Class Is In Session! Welcome To Cap Rate 101

Alright class. Today, we’re talking all about Capitalization Rate. And while this metrics got absolutely nothing to do with grammar, when it comes to its use in commercial real estate, you’ll most certainly want to mind your p’s and q’s.

Capitalization Rate, or Cap Rate as it’s known on the street 😉, is used to calculate the annual return on a property and gives insight into the value of that property. For an investor, it helps them easily compare multiple properties to determine the best investment. It can also provide insight into the fair market value of a given property when looking to sell. On the lending side of things, Cap Rate can tell us about the level of risk associated with that property and can be used to measure and manage those potential risks. Tracking the Cap Rate over the lifetime of a loan can help lenders determine how much they can expect to collect back from a property ­— which is important, especially in the instance that the borrower defaults on their loan. Yikes.

To calculate the Cap Rate of a property you’ll need to know the Net Operating Income (read more about that here) and the overall Property Value:

While this formula might not look very intimidating, I promise you, there’s a lot more here than meets the naked eye. Let’s walk through an example: say our property has a sales price of $1M dollars and an NOI of $100k. In that case, you’d get a capitalization rate of 10%.

What does that mean, exactly? As I mentioned before, the Cap Rate is the amount of return you can expect to get on a given property. That means the higher the cap rate the better the deal, right? Not so fast. The Capitalization Rate is also tied to the risk that’s associated with that particular property, which goes to say that it’s got an inverse relationship to the property value.

So typically, if you have a higher Cap Rate, you can expect to see a lower property value, and vice versa. In commercial real estate, high Cap Rates might mean that the property is located in a less than desirable neighborhood, or has a lot of maintenance fees due to the condition or the age of the building. If you have a property that’s located in a “low Cap Rate” area that might mean it’s in a great neighborhood, it’s a new building, or that it has a lot of appreciation potential.

Put more simply: the riskier a property is considered to be, the higher you’d want the Cap Rate to be. That ensures you’re getting a higher return on your investment to offset the higher risk associated with the property. And this checks out. Sure, you might be receiving a higher return, but you could also be spending a whole heck of a lot more on repairs for the building, or higher vacancy percentages due to unstable tenants. Conversely, the value of a property goes up when a property has a reliable tenant and/or is in great condition because it has a steadier income stream. There is less risk/work that goes into collecting rent on time or maintaining the condition of the building.

Generally speaking, anything between 4-10% is considered a good cap rate, however, this range is subjective because it can vary depending on someone’s investment or lending strategy.

Interpreting Cap Rate

While Cap Rate can tell us a lot about the value of the property, the associated risks, and the overall rate of return on investment, it’s not the only indicator of an investment’s strength. As a matter of fact, it falls short in that it doesn’t consider irregular operating costs/expenses or any value-adding variables. What’s more, is that the market Cap Rate can vary based off different factors like asset type, geographic area, submarket, and property class.  So, while the formula itself might be relatively simple, interpreting this metric is heavily nuanced.

It’s also highly sensitive to several different factors: Current market condition, current lease lengths, and in-place rents versus market rents can all affect the Cap Rate. The current market plays a large role in the property value which significantly affects the Cap Rate. In a down market, property prices drop causing Cap Rates to increase. In a high market, property values increase, and Cap Rates decrease due to the competitive nature of the market. The Cap Rate is also affected by rents because it is based off a property‘s in-place rent revenue. However, if a building has existing rents that are below market value, then the operator can increase the rent over time to better reflect the current market rents. This will increase the NOI of a property and therefore decrease the Cap Rate.

But wait, there’s more! Lease expirations can have a major effect on the risk of a CRE property — especially in a single tenant building. This is because the owner is relying on the rent revenue to cover expenses and make profits. In CRE it takes time and money to find new and reliable tenants to lease space. So, if you have a building that has a large percentage of expiring leases, then it is considered a riskier property which can create a higher Cap Rate. Not so simple anymore. See what I mean?

Here at Blooma, we pull in an extensive amount of market data and can provide the current Cap Rate for a property while taking many varying factors into consideration. Blooma also takes it a step further by then providing an income valuation based off the property’s current performance AND future proforma projections. This allows the investor and lender to see how profitable a property currently is and/or how profitable it is projected to be in the future. Blooma pulls in a ton of market data so we can provide our customers a Cap Rate that’s based on the asset type and the region that’s associated with their property. All that goes to say: when it comes to Cap Rate, we’ve got your back. No guesswork required.

And there you have it: Cap Rates 101. You passed 🤓 Any further questions? Feel free to stop by my office hours, or drop me a note at laurabohlmann@blooma.ai.

Until next time friends.

Bank FinTech Showcase Interview Series

Robert Albertson: Joining us is Shayne Skaff, Co-founder and CEO of Blooma. Also joining this interview is Carson Lappetito, CEO of Sunwest Bank a $2.2B asset bank based in Irvine, CA operating across four Western states.

RA: Shayne – Let’s start with you. Thank you for this opportunity. Let’s begin with your background and how you came to develop Blooma. Where did you begin your career, and why did you wind up in the commercial real estate space?

Shayne Skaff: Thanks for having me. I’m a career tech guy. So, I’ve spent most of my career building AI and ML platforms. My last company was a company called Maintenance Net. We automated how companies manage and transacted annuities. Mainly selling into kind of the service contract renewal and software license renewals space. I didn’t play in or really wasn’t around the lending or banking industry at

all until January of 2018. My co-founder, a gentleman by the name of Mike Persall, is the chairman and founder of a bank called ABP Capital, out of Encinitas, California.

Mike and I have been friends for a number of years. In January of ’18 we began looking for insight into what was going on in and around the commercial lending space from a tech perspective. I had not been tracking that space. Over a series of conversations he got me interested enough to dig in a little bit. He was looking for a solution for his own banking needs. He wanted to automate how they originated loans and went through pre-flight within the bank on their commercial real estate loans, as well as automate how they monitor their current loan portfolio.

I looked at the solutions out there in the marketplace where there are gating factors that got me into this. When I looked at what was happening on the residential side, it seemed that commercial lending was at least 10 years behind from an innovation perspective. Yet, looking at those customers that were buying innovation on the residential side, they had massive commercial businesses that were just largely going unaddressed. That, for a technology sales guy, was huge. It told me there was just a lack of product in the marketplace. In June of 2018, Mike and I launched Blooma.

I brought over some of my old team members from my last company, specifically on the product side. Mike gave us access to all of their underwriters within the bank. We mapped out all of the manual processes that went into originating new loans. Then I asked my product and development team what can the machine do here to move these individuals to the art of their job? That was the summer of 2018. We started to build the product.

RA: You hit the bull’s eye in terms of banking, because commercial real estate is usually the largest asset class in any community or regional bank. Can you take a moment and describe the Blooma platform suite and make the case for data-driven analysis and decision making in the underwriting exercise? And then attaching an impressive dashboard for quality and performance indicators, all while reducing the paperwork stream and ultimately unnecessary labor costs? How well has this been working in real life since three years ago, compared to the more labor intensive model?

SS: When you look at the space, what you generally have everywhere from the national, even global banks down to the community banks, is lots of point solutions. You’ve got solutions that do OCR, you’ve got cash flow analyzer solutions, you’ve got your Excel spreadsheet models, you’ve got third party databases, like CoStar and REIS and such. None of them are really connected. What you end up having are humans bringing all of these disparate systems together, which really causes the massive time sink. We’ve bundled up all of these point solutions into our Blooma platform, our machine.

On the front end Blooma allows commercial banks, and really any CRE lender, the ability to actually create lending profiles. I can create profiles on Blooma like a Fannie profile, a Freddie profile, or maybe a regional lending profile. From aggressive to conservative. Those profiles have thousands of attributes that make up a lending profile, but most of our customers really only look at 10 to 15 items that really move their

lending decision. They can tell the machine basically what’s important to the bank within those profiles.

The machine learns what’s important to that lender.

On the side of OCR, we can read offer memorandums, broker packages, P&L unit mix, rent roll and the like so the machine can automate the way they’re digitized. We’re connected to a number of third-party databases, where we’re actually pulling in things like rent comparable sales, environmental data and tax assessor data. That’s all coming into the platform. We’re doing all of the cash flow analysis, all of the valuations in Blooma. On the borrower/guarantor side, we have the ability to read tax returns, K-1s, bank statements, do all the spreading.

At the end of that process, we have the capability of providing a lending score, which tells our client the probability that the deal meets one or multiple profiles which they’ve set up in the system. They can go in and interact with that analysis. Even if they’ve got an LOS system in place we can actually deliver all of that data back into a loan origination system that they’re using or even more importantly, Excel models. We can actually populate their own Excel models with the resulting analysis that we come up with within Blooma. For our customers, they experienced anywhere from an 80% to 85% reduction in man hours on the pre-flight side. More importantly, what they get is the increased production on the business side. Now the client can do a lot more with the same workforce that they have in place because they now have a machine that can help reduce those man hours.

RA: You’ve developed a pretty broad-blanket, technological solution for data analysis. I think a lot of real estate professionals are concerned with the paucity of quantitatively comparable information that could be very useful for the lending decision. Why do you think this lack of data developed in the first place? Analysts in the industry often wail about the repeatability of real estate credit loss cycles that seem to only become obvious well after the fact. You seem to have addressed this.

SS: We certainly are trying. The data side is probably the most complex piece of this. There are a number of different data systems out there. I would say that some of them match up and some of them don’t. So the complexity is in how you gather all of this data, how you normalize that data, and how you bring together a better, more complete data set that you can work off of. That’s where I’ve spent my career, in a very nerdy place.

These are the things that we know how to do really, really well. We look to build up our data library, and find the best pieces of data in very specific places. For instance, via asset class. There are some third party data systems that have better information on say, multifamily, than other third party data systems. We go looking for and finding the best-in-class data in very specific spaces. So, when it comes together in Blooma, our customers are experiencing a very complete and accurate data set that they can work from. When you get better rent comparables and better sales comparables, you get better valuations. That’s really what we tried to produce out of Blooma.

RA: I take it you can do this pretty much agnostic to geography. Are there any restrictions?

SS: Only by geography within the US. We are US only today. We certainly have plans to move out but right now we’re US only.

RA: So, is Blooma’s secret sauce basically the ability for artificial intelligence to synthesize various sources of data into hard, usable data for underwriting decisions?

SS: I’d say there are a couple of pieces where we have extreme IP on the AI side. We’re not using AI to make decisions for people. We’re really using AI in two main places. One is in the parsing of data. Two, is when users are interacting with things like rent comparables, sales comparables. They’re actually training the system on what’s important for each of those asset classes that they’re looking at. On the data side, as we talked about, how we aggregate and bring this data together, and normalize it, we’re doing some really interesting things there. We have a product called Blooma Central, where all of that data structure is coming together. I’d say that piece is as impactful and as important as what we’re doing on the artificial intelligence side.

RA: Artificial Intelligence is increasingly ubiquitous in FinTech, from what I can see. Yet, the artificial intelligence concept is probably not comfortably or sufficiently understood by your average banker. Can you take us into the machine learning weeds for an example, perhaps, that bears on your applications?

SS: The uncomfortable feeling that people everywhere – not just in banking – have with artificial intelligence or machine learning, is really around the replacement of people. We don’t look at that at all when we build or when I have ever been involved with building out an AI platform. It’s really about moving people to the art of their job and actually making their job better. Utilizing the machine to do a lot of the monotonous boring work that a person doesn’t need to do, and really move them to the place where people are needed. One specific example would be around new construction lending: There’s more nuance in new construction lending than in some other asset classes, like vanilla multifamily loans. Where we can actually provide freedom for people to move to the art of their jobs is needed within underwriting and analytics. That’s really what our goal is. It’s about providing results that a person could use to make quicker, better decisions off of, instead of having the machine make the decision for them.

RA: Tell us what your client list looks like at this juncture. Is there a range by size or geography or type? Do you have pilot programs for the larger banks in the system? How many clients are on the Blooma platform now?

SS: We have a dozen clients on the platform at present and they range anywhere from the $1B to upwards of $10B. I’d say most of them are between $3 to $10 billion in CRE AUM at this point.

RA: Any pushback when you’re trying to put this together? In your view is there a typical profile of a bank to make the most of the Blooma platform or is it pretty broad?

SS: I wouldn’t even consider any feedback to be pushback. We’ve gotten such an amazing positive response. I’d say that really the time it takes to get to a yes is primarily because there’s so much involved

with vendor onboarding. This is more on the commercial bank side. On the private bank side there’s less of that that we have to go through. So, I’d say that our hurdle on the commercial banking side is really regulation and going through the vendor onboarding process. Going back to where the art of lending lies. There’s a lot more art on the $100M+ loans. I’d say our sweet spot sits between the $5M to $100M mark. I say that only because that’s most of what we have in our system today. It seems to be where we’re seeing the most activity within the Blooma platform.

RA: I assume you’ve had contact with bank regulators. What are their attitudes about your platform and their general receptivity? Have you brought them into the tent?

SS: It’s a great question. We haven’t had direct contact with bank regulators. Through the clients there’s been that contact, but not directly with Blooma. We’ve had no pushback so far. We’ve moved through vendor onboarding, in which case there’ve been regulators that have seen what we’re doing. We’ve had no real obstacles that have held us back from onboarding a customer to date.

RA: How do you price Blooma to the client?

SS: Today, our pricing is based on the AUM that’s actually flowing through our system. So, it’s not user based. We want as many users in our system as possible within our client base. It’s a fixed subscription fee, based on AUM. It’s pretty simple.

RA: Let’s move to the profit and loss impact. I know you say there’s about 80%+ reduction in costs. What does that do for a CRE-oriented bank’s bottom line? Can you turn it into dollars or some sort of impact on profitability?

SS: That’s probably a better question for Carson. From an ROI perspective the 80 to 85% is on man hours. I think because our costs in general come in somewhere around a basis point on AUM, we’re fairly inexpensive for the value that we’re providing our client. We’ve had no real pushback on our pricing model. From a bottom-line perspective, we’ve heard less, and again, Carson, it’d be great to get your comments on this. We’ve heard less interest on the cost take out and way more interest on the production increase. They’re not letting people go by using Blooma. They’re just doing a whole lot more with less. Carson does that resonate with you?

Carson Lappetito: Absolutely, I can add on two ways. One being on the loan origination side, I think about it the same way I think about my salesforce cost. It’s very easy to do the math of what the Blooma platform is costing you and put it into the spread you’re originating off a loan. You divide the cost of Blooma divided in by your 2% spread, dropping to the pretax income line item. You say am I going to drive, incrementally, that many more loans, by having this platform? When you do that math on a platform like Blooma, that increases your efficiency and your ability to deliver term sheets and lending at speed, which is ultimately how we differentiate ourselves in the relationship and the commercial real estate lending business. It’s a no brainer decision.

RA: Thank you, Carson. Tell us how long you’ve been on the platform and what was it like to implement it? How are the results versus your initial expectations? What’s different now?

CL: I’ll caveat this answer with I’m different than most bankers. I have approached Blooma a bit differently than a lot of other bankers. My answer is not indicative of all of Shayne’s clients. I was very focused on leveraging Blooma as a portfolio management tool, less so on the front end origination to start. We’re running our pre-flight through that process, in order to get to a quicker decision on putting a term sheet out to a client.

We’ve been engaged for just about three months now. We have our entire portfolio loaded up. A big chunk of our rent rolls, a big chunk of our appraisals have been OCR’d through their AI into the platform. Basically, we can look at our entire loan portfolio with real time comps. With real time cap rates from transactions generating real time values of what the portfolio looks like, both as a trend line from what the appraisal was when we originated it. What is the Blooma value today based on their public comps? Especially when we load in our rent rolls. Then we have a rating score of how we feel about the current rating compared to where we originated it. How does it stack in our aggregate portfolio? Should we be spending more time on this credit or less time, depending on how that ranking looks.

What we’re really driving here is the fact that banks, historically, go and look at how their portfolio performs on two occasions: one during the annual review, or two, when a payment is missed. They don’t do any portfolio management in between. In real estate, usually, you have a one-time check on a real estate loan. If you go back to the real estate world. You really aren’t tracking what’s happening in that local economy, that intersection, etc. on an interim basis between your annual reviews, and so especially in an environment like we’re in right now where we’re going through COVID, different asset classes are performing differently. Traffic patterns have changed.

You don’t necessarily know what the exact per square foot rental comp is happening on that corner, that sub market of the city. Blooma is pulling that data real time and telling you where you need to focus. So, for us, it puts us in the catbird seat of looking at our portfolio and saying, “What are we very comfortable with?”, “What are the emerging trends we have to be aware of?” It allows us to be really proactive in managing the portfolio as opposed to being reactive, which is how almost the entire banking industry manages their loan portfolio.

RA: So, is it fair to say that you feel a lot more comfortable with your credit exposure and your portfolio than you did before? Is it essentially no longer an annual review issue? It’s a real time, ongoing review.

CL: Yes, I feel a lot more comfortable. I’m an incredibly conservative underwriter. I would say the proper term would be that I feel a lot more knowledgeable. Therefore, the deals I feel uncomfortable about I’m on top of and dealing with, and the deals I feel comfortable about my team’s not spending time on. They’re spending time either managing the deals we’re worried about, or finding new business, which is ultimately how you pay for the whole platform.

RA: Okay, a question for you both. What is Blooma’s differentiation to the competition? Carson, you go first please. Why did you pick Blooma, and what advice or argument would you give to prospective clients?

CL: I don’t know of another platform in the space that is doing this. So, the decision really is, am I going to do it the way I’ve always done it, or am I going to implement Blooma? I think we’re on the edge of more companies and more technology entering our banking space to help make our whole industry more efficient. Today, if you want to do what I’m talking about here when Shayne was talking about either on the front end, or the portfolio management side, there are no other options.

SS: I’ll just add to that, Robert. We’re the only ones actually taking all of these point solutions out there and bringing them together. I’d say if you take any of those point solutions, there are all kinds of competition going on within them. I would say we’re not really competing against any of them, even the third-party data providers that we don’t have contracts with. I don’t look at them as competition. The reality is we may decide to make them part of our library six months from now because we like the data that they have. Right now, we’re in a pretty good position from a competitive perspective. There’s a lot of big players out there. I generally am looking at what they’re doing. It’s mostly the guy or the girl that’s sitting in a dorm room right now who’s building new crazy products that are going to come up that we’re going to see. I think we’re going to see a lot happening over the next couple of years as PropTech is really moving quickly now. Since 2018, when we started Blooma, I’m starting to see a lot more activity happening within the PropTech space specific to commercial.

RA: So, is it fair to say you’re unique at the moment? You’re the full package?

SS: Yes, I’d say so. I haven’t seen somebody that we’re competing head to head with that has the same solution as Blooma. There’s lots of companies that have pieces of our solution, but we’re kind of packaging up lots of these point solutions in a very unique way. I’ll add it’s not just the product build itself, but also how we’re implementing the product within our customer base. We recognized really early on the complexity of the commercial banks and the core systems that they had in place. We are very sensitive to not having to forklift out one system to put in our system. What we’ve done is very unique.

We’ve built a very flexible platform that can be integrated with, for instance, nCino, or any other LOS system that can feed your Excel model that you might want. Again, the results of our data are going into some other system that the bank might have. On the preflight side you can literally access our system in a few days. We can have a system up and running for a bank in a few days. They could be going in and loading OMs and actually doing deals in our system with very little effort and no implementation. The only time sink that we have is if we’re integrating into a system that the bank has that we haven’t seen before. Generally, if there’s an open API, it’s a very quick build to get us connected.

RA: You’re basically core agnostic?

SS: Yes.

RA: Let’s move over to the auditing and monitoring function of Blooma. How does that work?

SS: The portfolio monitoring? So basically, I walked you through the whole analysis when we’re doing the initial cash flow analysis and the valuation of that asset. When we consume a customer’s portfolio, we’re literally doing that every single day on every asset in that portfolio. What does that mean? Does it mean that they’re going to see a change in the lending score every day? No, because the times when there are changes are generally when there’s new data to be presented to the machine for the analysis. If that data changes, like an updated rent roll, if it changes to the positive, then it’s generally going to move the lending score up. If it changes to the negative, it’s going to move the lending score down. If there’s some dramatic increase in vacancy rates in a specific region and it impacts some of the assets that a bank might have there, then that will drive the lending score down. That’s happening every day. Our system is looking for the changes that might exist with that asset, or if they upload new financials for the borrower that could also have an impact on the monitoring of that specific loan.

RA: Carson, you’re unique as well, in that you cover a very large swath of geography in the west with a very thin branch network. Are there any examples so far of how this function has been turning up things that you appreciate in terms of warning signs or something you might otherwise have missed?

CL: There’ve been a lot of great examples. The first exercise our team did was a COVID analysis on all of our loans. We ranked them by risk level and in various tiers. We took that and compared it to the risk scoring that came out of Blooma. We looked at the differences and asked “Why?”. There were a number of loans on that list that we thought were better than they actually were. We’re very strong underwriters, but the goal is to figure out what you don’t know so you can solve it, and there were a handful of loans that had higher LTV than we had thought based on rental rate compression in their market. As a result, they’re a bit higher risk than you initially anticipated. Then you can adjust accordingly when you have your covenant testing periods.

You want to make sure you’re on top of those from a portfolio management perspective. Banking’s a risk- based model. You can’t be a hawkeye on everything when you have a big portfolio. You have to make sure you’re on top of everything that’s got a little more risk to it. Just the fact that it is teasing out something that you should be looking at and directing an RM towards is very valuable. Helping that RM focus on the art of their job.

RA: I will compliment you on your speed of implementation. It is unusually short for within the FinTech arena, which I think is quite impressive.

Let’s finish off with a question to Shayne on scope and scalability of your organization. He recently received an injection of capital from Canapi. How comfortable are you with your funding to date and your scalability going forward? Specifically, where do you think your client base will be in another year or two?

SS: Very comfortable with the capital that we have right now. We have a very strong balance sheet and will need it to do the things that I’m forecasting to do over the next 18 to 24 months. We’re using that

capital to bring on a sales and marketing organization. We’ve been heavily driven by R&D and product. Bringing on sales and marketing is something that I started doing at the beginning of this year. We brought in a head of sales, we brought in a head of marketing, we brought in sales directors, we’ve increased our customer experience (onboarding) team. They’re the ones that bring our customers on and train them.

It’s a pretty easy onboarding because we’ve got a platform that’s fairly intuitive. Users only really require an hour or two of training. The Canapi guys have been phenomenal to us. I believe in raising capital where capital is the least of what you get. It’s really the intellectual knowledge and the help on the other side. That’s when you know you’ve got a great partner, and that’s what we’ve got with Canapi. They’ve given us great introductions. Not just into clients, but into resources which are extremely important for a company of our size.

We’re looking to basically triple the quantity of customers over the next 12 months. I feel very, very good about doing that. I’ve got a great operator in place that understands what we need to get to certain places. He was my COO at my last company. So, we’ve been together a long time. I’ve also operated with my CTO before. This is the team that I have in place – the core team that has been here since the beginning. We’ve all actually operated together and that makes for much easier momentum. I tell them what we’re doing from a sales and marketing perspective. They know what I need from a product and R&D perspective to make that happen.

RA: It sounds to me like the Blooma story has a lot of impact, and Canapi is a great partner. They’re best in class in terms of industry knowledge and connections.

Have I missed anything, anything either of you want to mention?

SS: I think you’ve done a great job, Robert, of hitting some key points that I would want to put out there. I want to thank Carson for his time for being a part of this. It’s always wonderful to have a client doing these interviews with me. So, thanks, Carson for spending the time especially what it sounds like in a very crowded environment that you’re in out there.

CL: Happy Shane brought a product like this to market. The more innovation we can drive in our industry, the better. I’m a happy client, and thanks for including me in the interview.

SS: I’ll just say, Robert, for those banks that are going to view this, the process is probably abnormal for them. I think this specific space has been stuck in a software environment that has been a super heavy lift for many decades. This is not that. That’s very important for them to understand. Initially, when we come in and start talking to a bank they don’t understand until we get through our meeting and they actually understand that this is a technology that’s built in a very different way than the system that they’re running on today. I look forward to talking to a lot of those banks over the next year.

RA: This has been a very helpful, thoughtful interview. I appreciate your time. Both of you. Shayne, if a financial institution reading this would like more information or maybe connect with you, would you be willing to provide your phone and email coordinates here?

SS: Yes. You can email me at ShayneSkaff@blooma.ai. I also can be reached by phone at 858-442-3696. I am happy to talk to anybody interested.

General Information and Disclaimers

This report has been prepared and issued by Piper Sandler & Co., a registered broker-dealer and a member of the Financial Industry Regulatory Authority, Inc. The information contained in this report (except information regarding Piper Sandler & Co. and its affiliates) was obtained from various sources that we believe to be reliable, but we do not guarantee its accuracy or completeness. Additional information is available upon request. The information and opinions contained in this report speak only as of the date of this report and are subject to change without notice. Contact information for Piper Sandler & Co. and the author of this report is available at www.PiperSandler.com

This report has been prepared and circulated for general information only and presents the author’s views of general market and economic conditions and specific industries and/or sectors. This report is not intended to and does not provide a recommendation with respect to any security. This report does not take into account the financial position or particular needs or investment objectives of any individual or entity. The investment strategies, if any, discussed in this report may not be suitable for all investors. Investors must make their own determinations of the appropriateness of an investment strategy and an investment in any particular securities based upon the legal, tax and accounting considerations applicable to such investors and their own investment objective. Investors are cautioned that statements regarding future prospects may not be realized and that past performance is not necessarily indicative of future performance.

This report does not constitute an offer, or a solicitation of an offer, to buy or sell any securities or other financial instruments, including any securities mentioned in this report. Nothing in this report constitutes or should be construed to be accounting, tax, investment or legal advice.

Neither this report, nor any portion thereof, may be reproduced or redistributed by any person for any purpose without the written consent of Piper Sandler & Co.

© 2021 Piper Sandler & Co. All rights reserved.

We Need To Talk…About DSCR

“We need to talk”.

I know they’re not words you want to hear, but I promise this time it’s worth your while.

About what exactly? Nope, you’re not getting fired, and we’re not breaking up. Today, we’re talking about DSCR — or debt service coverage ratio and why it’s basically the Holy Grail of CRE metrics.

DSCR is a calculation used to see the cashflow of your property after you’ve removed all of your annual debts, but what you really need to know is that it’s important. Like really important, and it’s generally the first thing lenders look at in the borrowing process. Why? Because it can determine the structure of a loan. For example, if someone isn’t even close to the minimum DSCR they are looking for, then the lender will decline the loan right away. If someone has a DSCR that comes very close to the minimum DSCR amount, then a lender might cut back the total loan amount which will then require a larger down payment. This will increase the DSCR amount AND lower the LTV amount which will mitigate the risk in two ways. See what I mean? Important.

How Do You Calculate It?

To calculate the DSCR of a given property, you’ll take your Net Operating Income (NOI) (more on that here) and divide it by the total annual debt.

Typically, a CRE loan has a business listed as the primary borrower. So, in this case the annual debt includes all of the current year’s business debt obligations (or borrowers’ debts). This would include debts such as line of credit payments, minimum credit card payments, business lease payments or any type of business loan payment (this can even include business equipment loans). The most common mistake I see in calculating DSCR is not including all of the borrower’s debts. Some people only account for the loan payment but forget to include all other business debts, so be sure to keep an eye out for this.

Let’s Practice!

In this case, let’s say you’re looking at a property with an NOI of $150,000 dollars and there’s a $10,000 dollar monthly payment. Over the course of the year (annual, or x12), that monthly payment would amount to a total Annual Debt of $120,000.

Divide the NOI by the total annual debt and that gives you a Debt Service Coverage Ratio of 1.25. So, what does that mean exactly? A DSCR of 1.25means that you have 1.25x the amount of cashflow on that property than you do debt (FYI, that’s good).

It’s important to note that preferences and requirements on what makes a ‘good’ or ‘bad’ DSCR value can vary from lender to lender. Generally, a lender will create a covenant saying that the property must have a “minimum DSCR of X” and then they track it through the life of the loan to ensure that the property is still making an adequate amount of money to make the loan payments. If the borrower doesn’t meet the minimum DSCR requirement at any point during review (typically tracked once per year at annual review) then the lender might start reviewing that loan more frequently (like every quarter, for example). Or, if this trend continues for long enough, they may possibly force the borrower to pay off the loan entirely.

While there is no ‘set range’, here’s generally how DSCR gets broken down:

No matter how ‘good’ the DSCR is at the outset, it always gets tracked throughout the year and here’s why: it fluctuates, and sometimes, a lot.

The DSCR value can change because the NOI can fluctuate, too, based on changes in things like vacancies and expenses. So, if your NOI decreases then your DSCR will decrease as well. Also, the total debts can increase as the borrower takes on more debt. If the borrower’s debts increase, then the DSCR will decrease.

Here Is What A Borrower Can Do To Work On Increasing Their DSCR:

  1. Increase their NOI by increasing their revenue (which they can do by raising the rent, for example).
  2. They can also decrease their expenses to increase NOI. For example, they can analyze their spending and cut down any unnecessary expenses.
  3. Lastly, they can pay off business debts. By paying off as much debt as possible it will decrease their overall total debt expense and in turn increase their DSCR.

We’ve established that DSCR is important, but today, it matters more than ever.

With market changes brought on the by the pandemic, we’re looking at a level of instability that makes lenders wary — and for good reason. In most cases (and pre-COVID), a ‘good’ DSCR value would be anywhere between 1.25 and 1.5. Today, you can expect that lenders are likely looking for DSCRs in a higher range than usual because of that volatility in the market. Raising the bar on DSCR helps lenders mitigate risk in uncertain times (like now) where expense rates and vacancies are higher than normal. As such, it’s extremely important for lenders to be able to consistently monitor the DSCR of a property with real-time data. This way, they can get the most accurate picture of performance possible and get ahead of unfavorable changes that may be coming.

And there you have it: DSCR. Hopefully this gives you a good overview of why it matters, how it’s calculated, and how to stay on top of it. Your takeaway for today? Keep your blood pressure down and your DSCR up! Doctor’s orders!

Until next time,

Laura

Still have questions? Drop me a note: laurabohlmann@blooma.ai.

Let’s Make Some (NOI)se For NOI

NOI. This powerhouse metric is a staple in CRE lending. And today, we’re breaking it down to the basics for you.

  • What is it?
  • How do you calculate it?
  • Why does it matter?

These are some of the questions I often get asked on the job – and for good reason. NOI, which stands for Net Operating Income is a critical measure of whether an income generating property is going to be profitable or not. Which, if you’re reading this and are in the world of commercial real estate, you definitely care about.

NOI measures a property’s ability to produce income based on the income from its operation and is often confused with cash flow. But make no mistake, these two things are not the same. NOI is different than cash flow because cash flow includes debt service for the property (debt service = debt payments). Cash flow is the income generated after the debt is taken out. Basically, that means the cash that the owner makes off the property after all the expenses are factored out.

If you want to understand what the NOI of a property you’re considering is, you’ll need to know how to calculate it:

First, we need to determine the Total Revenue of the property. So, for example, let’s say you have an investment property that makes $15,000 per month in rental income, after you subtract out any of the vacancies and concessions (more on that in a moment), you’ll multiply that by 12 – which puts you at $180,000 per year (your Total Revenue).

Next, you’ll need to subtract out any of the operating expenses. What falls into that category? Things like repair fees, maintenance fees, property management fees, insurance, etc. Now, I know what you’re probably thinking: where’s the mortgage payment in that list? Surely that would be included in your list of expenses for the property, right? Not exactly.

Technically, we don’t consider your mortgage as an operating expense, but rather, what’s known as a ‘finance expense’. Here’s why that matters: finance expenses are incurred by the owner/investor of the property rather than the property itself. (This is important to note because NOI is only intended to capture the income produced by that property. Finance expenses, on the other hand, can vary from borrower to borrower depending on how much they are financing their loan for and the structure of that specific loan. In other words: It’s not relevant in this case.) So, for this example, let’s say you have $50,000 of operating expenses each year. You’ll subtract that from your $180,000 which leaves you with $130,000. Your NOI.

The math is simple enough, but the result matters. A lot. Not only is NOI a really simple and reliable way to get at the value and potential of an income producing property – it’s also used to calculate lots of other critical CRE metrics as well. For example, you’ll use NOI to determine things like:

  • DSCR (NOI / Annual Mortgage Debt)
  • Cap Rate (NOI / Current Market Value)
  • Property Value (NOI / Cap Rate)

A Brief Interlude on Vacancies and Concessions:

Let’s get back to vacancies and concessions for a moment. Vacancies take into account the % of rental space at a property that is not currently leased, or is currently vacant. We remove vacancies out of the calculation because we currently aren’t collecting any rental revenue from that space. Concessions are typically given to someone as an incentive to sign a lease. Typically, this would be in form of free rent for a new tenant. For example, many places will give you one month of free rent if you ‘sign today’. In cases like these, you’ll want to make sure that the one month of free rent is not included in total revenue because it was not technically collected.

…Aaaaand that’s a wrap on NOI! Hopefully this gives you a better understanding of how it’s used and its importance to CRE lending. Still have questions? Drop me a note: laurabohlmann@blooma.ai.

Until next time!

Connect with Laura on LinkedIn

The Power of Curated and Consolidated Information

Answer to any question: 10 cents. Correct answer to any question: 10 dollars.

The information revolution has brought access to data to the masses. “Knowledge is power” is not specific enough anymore – the true challenge is in making sure the information used in decision making processes is reliable.

If you want to purchase a product, you spend time scouring the web, learning about the product’s characteristics, comparing it to its competitors, finding different suppliers and their pricing, etc. In order to make an informed decision you know that you must become a mini expert on that specific product. Businesses make this process more efficient by building information harvesting systems which automate a portion of the research, but these still require employees to invest their most valuable asset – time. Time to make sense of the information, to make sure that it is relevant to specific business needs, etc.

First, every piece of information has a finite lifespan. When you check the weather, you know that the information you just gathered is good for today only. The value of a commercial real estate property fluctuates as market conditions change, and a year-old credit score may not reflect the current financial status of a borrower. Therefore, it is important to identify the age, rate of accuracy degradation and expected expiration of each piece of data that you harvest.

A successful information harvesting and processing system is characterized by the user spending little to no time gathering and processing information and most of their time making decisions.

Next, you need to identify the credibility of the source of the data. Did you get the borrower’s net worth valuation from the borrower or a third party impartial source? Try reading about the same event at two different news outlets. The same story looks totally different if you experience it through different sources. You may have to read multiple sources and find the truth yourself somewhere in between them. This is fine when objective truth is not so important, but when information needs to be consumed in a path that leads to the success of the person or business entity, truthful data is crucial.

The situation gets even more complex when we consider the concept of infobesity. Information obesity is just what it sounds like – an epidemic that exposes humans to an overload of information, leading to a situation where we can’t see the forest for the trees. When an underwriter is onboarding a loan, they put a lot of effort into harvesting huge amounts of information, and at some point they need to make a decision based on a forest of data points. It’s like telling your doctor that you are not feeling well and explaining your symptoms, and in response he or she simply gives you a list of the hundreds of possible causes for those symptoms rather than telling you their qualified diagnosis. That’s the experience we are used to today when we use search engines, and it’s very similar to what decision makers face when they need to decide based on thousands of data points around a loan.

So, in our era, rather than saying “knowledge is power,” it is more precise to say that “curated and consolidated information is power.” The most valuable information must be curated by an expert (human or AI) to fit the needs of the business, automatically harvested by a computer system, and continuously checked for its lifespan, credibility and accuracy. And when all of that is done, the business must take the next step of consolidating the information in a form that will still deliver the powerful insights without causing an outbreak of infobesity.

Today, decision makers spend most of their time preparing and processing information and only a small fraction of it making business crucial decisions. A successful information harvesting and processing system is characterized by the user spending little to no time gathering and processing information and most of their time making decisions. Such users can conduct much more business in the same amount of time, which leads to a higher chance of business success and a lower “cost of doing business.” Such a system allows its users to look at information like a judge on “America’s Got Talent” might look at a contestant. They can watch just a few minutes of a performance and make an informed final judgment without further investment.

Now that’s power.

How to Build a Disruptive Product for a Conservative and Highly Regulated Industry

Seven tips for success.

I’ve been designing, building, and commercializing enterprise software products in several industries for the past 20 years. When Blooma was founded, I faced the new challenge of building a disruptive commercial lending platform for one of the most conservative and regulated industries – banks and financial institutions. Looking back at the last 18 months, it was an exciting and enlightening journey. We are getting great feedback on our product, particularly on our ability to utilize artificial intelligence to automate data collection, calculations, and valuation models, while enabling lending professionals to use their experience and talent in the art of their work.

I have learned a lot along the way so wanted to share my 7 best practices for anyone planning to build and launch advanced technologies to conservative industries that may just be catching up on disruptive innovation.

1. Build the product with customers for customers. We developed our product together with to-be-users, by partnering with future customers and hiring professionals from their specific industry. This is especially critical when the product team lacks experience in the industry that they are building for. The combination of views from both inside the targeted industry (for better product fit) and outside of that industry (for the “art of the possible”) typically leads to better innovations.

2. Combine user-centered and vision-centered innovation principles. Start building your product by applying user-centered innovation methods. Strive to understand the needs of users through deep analysis of their behaviors and how they interact with existing products. Later, vision-centered innovation methods should be added. This is required in order to focus both on doing the same things better, and also on doing things differently. During our development process, I pushed the team to bring up ideas and concepts that customers did not specifically ask for or did not think were technically possible, and some of those ended up being key parts of our solution.

I have learned a lot along the way so wanted to share my 7 best practices for anyone planning to build and launch advanced technologies to conservative industries that may just be catching up on disruptive innovation.”

3. Build flexible workflows using multi-services architecture. Each customer has their own way of doing things and requires the technology to adapt to their own workflows. This is simple when you build a custom solution, but much harder with a SaaS platform that is used by many different customers, where typically “one size fits all.” Early on, we decided to build a flexible architecture, allowing users to configure their own scoring algorithms, workflows, and how they would like to consume the asset and borrower analysis (in our intuitive UI, a summary report, or even directly feeding their own Excel spreadsheet models).

4. Build your human/machine interaction to be mutually beneficial. No matter the level of sophistication of the artificial intelligence and automation that you develop, it must be clear that the human users are always in control. AI can enable customers to make scientifically informed decisions, but it should not make decisions for them. Our product provides ways for users to validate AI-driven information gathering and to easily correct it where needed. For example, we use machine learning models to select the best sales and rent comparables for an accurate valuation, but users can still include or exclude comparables based on their specific experience and knowledge of the area. Not only do users end up getting better data that exactly fits their needs, but these interactions help train our AI models and make them even better.

5. Pay attention to the unique needs of the sector you’re serving. If you’re working with AI, the type of models that you develop must fit the needs of the specific industry that you’re building for. In more conservative and highly regulated industries there will always be a need to know exactly how the machine came to the results that it did. You should avoid deep learning models which typically don’t provide reasons or explanations. Think of the machine learning-based movie recommendations in Netflix. Based on a list of movies that you’ve watched, they tell you that you’ll probably like a certain recommended movie. If Netflix recommended a movie for you without explaining why, you might be less likely to pay attention. And in a heavily regulated industry, a conclusion like that without an explanation would not be acceptable at all.

6. Keep it simple and avoid developing complex technology where it’s not needed. Technologists many times are so focused on the art of the possible that they build features that will never be used just because they’re cutting edge or the latest trend. I say, where advanced technology doesn’t make sense, don’t use it! For example, you can spend months building complicated screens and models that are expensive to build and hard to maintain, when what users really want is a simple download to an Excel spreadsheet. When I review each proposed new feature, I always consider usability, simplicity, and maintainability in equal measures.

7. Build the “leap” disruptive innovation in parallel. Sometimes customers (or industries) are just not ready for truly disruptive innovation. So, while adhering to the previous recommendations, I always maintain a product innovation track where we research and develop truly disruptive innovation that might not be accepted today but could be eased into in the future. For example, today users would probably not accept making loan decisions based on an AI model, but we can build this model in the background while still allowing users to analyze risk the way they are used to. Then sometime in the future when we can show them how accurate the AI model would have been in predicting a loan default, they might be ready to adopt it.

Hopefully you’ll find something helpful in these 7 tips. Introducing something new to a sector that has been satisfied with doing things the old way for a long time can be daunting. But it can also be fun, especially when you know that you’re truly making a difference.

The Art of Shaping the Story

CRE service providers and the current technology revolution.

When I bought and refinanced my home, I opted for a lender who provided an online experience, and it was pretty painless. It was clear to me during the process that the world of residential real estate has undergone a technological transformation, and I think most consumers have embraced that change or at least benefited from it whether they noticed or not (and maybe that is the best measure of an industry’s success at implementing technology).

Commercial real estate? We have some catching up to do. Why have we not embraced technology at the same pace as our residential peers? Well, like many things in life, it’s a numbers game. To oversimplify, our residential counterparts have large numbers in the amount of transactions and we tend to have large numbers in the dollar volume of transactions. Technology can do wonders for process and efficiency, and the residential world, with its long tail of high transaction volume, had more to gain from early investments in technology. Most in CRE believe that the personal touch of an old-fashioned phone call or in person lunch (pre-2020) carries with it a sort of cache. And I tend to agree, but technology is entering our industry at a rapid pace and it’s not replacing the old ways as much as allowing everyone to elevate their game (lenders and brokers included).

“No deal is the same, and every deal has a story. I work for a commercial mortgage brokerage, and our goal is to find the most efficient way to get a property’s (and borrower’s) story to the widest audience of lenders possible.”

There’s no doubt that CRE loans will never be as easily standardized as residential ones. I like to say that in the world of CRE the true value of a building really comes down to 3 things: what it costs to build, what the neighboring building sold for, and what the tenants are paying in rent. No deal is the same, and every deal has a story.

I work for a commercial mortgage brokerage, and our goal is to find the most efficient way to get a property’s (and borrower’s) story to the widest audience of lenders possible. Remember the pre-pandemic days when we all used to shop around for airfare to travel somewhere? You weren’t just looking for the cheapest price to fly. Based on departure dates/times, airlines and even airports, it’s likely that you decided on a flight based on more than just price. In our industry, there are a lot more variables than just destination and travel dates. A borrower may in fact be able to walk into a bank, advocate for their deal and get a loan quote, but a broker will give them a whole matrix of options with different features – the majority of which they wouldn’t have found on their own due to time constraints and access to capital providers.

Therefore, a broker’s job is to look at a loan and put it into a format that best tells its individual story. As we know, much of that work is done very manually today and involves pulling reports, formatting data, and constantly updating our analysis as new information arrives. Brokers also have to do the soft marketing, with analysts using design software to format everything into official presentations and loan packages. The point is, it’s very manual, and we’re often starting every story over each time with “once upon a time…” with a pen on a blank notepad.

But it doesn’t have to be that way. It’s a great time for the commercial side of the real estate industry to embrace the technological advancements of the last few years. Many of us have pushed PowerPoint templates, Excel macros and PDFs to their limits, after all. But I’m not just looking for an automated process for creating nice looking presentations – it’s a lot more than that. Integrations with data partners through API’s and financial spreading through document recognition truly are game changers in my mind. When the info, like comps, market narratives, and property data can be gathered and entered via machine, the analysts can focus on analyzing that data and shaping it into a compelling narrative.

In other words, an analyst’s job can move from pulling reports, entering data and formatting documents to truly analyzing deals. When I was working as an analyst this would have been a game changer, and my productivity would have multiplied (not to mention my job satisfaction). I’m sure my team would have loved the gains in productivity, and I know our clients would have loved rapid turnaround times. When it comes to technology, I think everyone can be a winner.

Still, technology isn’t going to revolutionize the industry overnight. Not every client wants to use their phone to get a CRE loan while in their pajamas, but that’s ok – I believe some will. In the meantime, my team is embracing the latest advancements in a lot of ways, but we still prioritize a personal phone call when it comes to getting deals done. It’s just that we can make a lot more of those phone calls and tell the individual stories of a lot more deals when technology has given us so much time back.

Industry Adaptation and Privacy in the New Era of AI

How data security is changing everything.

In a collaboration between BCG and MIT from 2018 (Ransbotham et al., 2018), researchers found that when it comes to artificial intelligence, companies and organizations can be classified into 4 groups: 18% Pioneers, 33% Innovators, 16% Experimenters and 34% Passives.

Pioneers are enterprises with an extensive understanding of AI tools and concepts, and embrace AI in significant ways. Innovators have a good understanding, but still display little actual application of AI in their business. Experimenters are using AI for their business, but without seeking an in-depth understanding of the AI methods. Finally, Passives lack both in-depth understanding and application of AI technology.

Interestingly, all four groups agree that AI will change their business model in the next few years. This means that sooner rather than later, AI applications will penetrate the entire corporate landscape. This will be important as AI becomes a pillar of competitiveness, allowing companies to get things done faster and more accurately and to reduce time spent on less desirable work.

“Private AI technology allows a company to use data to train highly performing models without exposing or sharing confidential data.”

In the past, laws around data privacy have hampered the evolution of AI applications in many industries which deal with private data, such as sensitive or personal information. Typical examples are the healthcare and financial sectors. In recent years however, emerging technologies have enabled the secure use of private data to build and use AI. These private AI technologies are accelerating the quantity and nature of novel AI applications, and transforming the AI landscape. Thus, Pioneers, Innovators, and Experimenters in the financial sector have started to use AI on their private data.

Private AI technology allows a company to use data to train highly performing models without exposing or sharing confidential data. Although the field of private AI is very cutting edge, there are a few notable methods already, including differential privacy, homomorphic encryption, federated learning, and data anonymization. For example, federated learning is an advanced, state-of-the-art method used by Google and the healthcare industry (full disclosure: Blooma uses it too). This method involves making use of client data, without the client having to share their data with anyone.

In a typical set up, a model is deployed to those client locations and trained on the data there. After training, the model is sent back to a central location where it is merged with the models of other clients. After this merge, you end up with a model which has learned from data across all clients without sharing the data in one central place. Federated learning, if done right, ensures that there is no exchange between each individual client’s data. Thus, federated learning allows a company to learn from confidential data while keeping the data secure and private.

Meanwhile, the introduction of AI into the corporate landscape has led to changes in legislation which continue to develop. The General Data Protection Regulation (GDPR) and California Consumer Privacy Act (CCPA) are two statutes that have recently come into play in the attempt to promote and regulate data privacy and security in the European Union and California, respectively. Additionally, Brazil passed the General Data Protection Law that came into effect in February 2020. According to Richard Koch, Managing Editor of the GDPR, EU, these legislations that have begun to emerge have some common principals, including the importance of defining personal data and certain fundamental rights relating to data subjects (2018). Private AI allows us to obey those laws while capitalizing on private data. Whether you consider yourself or your business closer to the Pioneer or the Passive end of things, your data is already a part of this changing landscape, and security will continue to be the driving factor in the future evolution of private AI technology. Despite the many differences between the statutes mentioned above, we can only assume this is just the beginning of the movement toward an increase in legislation to govern data privacy.


Philomena Lamoureux
Philomena Lamoureux

Philomena Lamoureux, Head of AI at Blooma, is a published PhD specialized in ML, NLP and privacy-preserving AI solutions.

The Life and Times of a Former CRE Underwriter

How the traditional approach can lead to more distractions than deals.

As Blooma’s Director of Customer Experience, I work a lot with lenders and brokers, especially their commercial real estate underwriting teams. I’d like to think I’m familiar with the motivations and struggles of our clients, since in a former life I used to be a CRE underwriter as well. The truth is, CRE underwriting and analysis processes are very fragmented, which leads to inefficiency.

Much of my job as an underwriter involved the management of many disparate data sources. For as large and mature as the CRE lending and investment space is, underwriting processes and innovation have not evolved or kept pace.

Got a new deal coming in? You’ll start by pulling up a model you used on a prior like-kind deal, and then overwriting the data. You then begin searching for sales and lease comparables on multiple sites, speaking with brokers about transactions and trends in the area, reviewing maps and street views to analyze the area surrounding the property, rounding up and reviewing title data, creating proforma cashflow models, the list goes on. Not to mention requesting and analyzing borrower financials, all in different mediums. You become a data aggregator, pulling in files, documents, and data from numerous sources and saving them all into a shared folder, drive, or system where they come together and start to make some sense. All of this work ultimately culminates in a final Excel model and a credit report/memo drafted in Microsoft Word. This is a painful process, and no matter how good you are at it, by its nature it is prone to errors.

“You become a data aggregator, pulling in files, documents, and data from numerous sources and saving them all into a shared folder, drive, or system where they come together and start to make some sense.”

An underwriter’s true job is to come to a binary decision as accurately as possible. Are we funding or passing on this? That’s it. But don’t get me wrong. Underwriters do not have a simple job even if their end goal is to make a decision one way or the other. And the “noise” of the process described above doesn’t help to get there – in fact I think it can mostly distract from this goal.

I think technologically we’re at a bit of a crossroads. Advancements in artificial intelligence have made it possible to take the more monotonous and time-consuming parts of the job of a CRE underwriter and hand a lot of this to computers. Tasks like parsing through borrower, financial spreading, searching for comps, etc. This allows underwriters to be much more efficient, and frankly frees them up to review each deal more analytically. To use their vast experience to engage in a strategic conversation about a deal and whether it should happen, which is what they do best.

If we hadn’t developed this approach at Blooma, someone else would have because its time has come. The only question now is how quickly lenders and brokers will embrace it. I’m excited working with all of the lenders and brokers that already have. It feels good to make the jobs of CRE professionals a little more enjoyable, by helping transition them to a newer way of doing things. It wasn’t that long ago that I was one of them.

Your Computer Gives You an A for Effort Today

Using AI to help you make decisions, not make decisions for you.

I demonstrate our product to many people, so I get to hear a lot of feedback and questions. One of the most potentially confusing aspects of what we do is using artificial intelligence and machine learning to generate a “score” for a certain commercial real estate property or proposed deal. It’s sometimes hard to explain what this really means, so I’ll try again here.

People sometimes think of the scores that are awarded an Olympic gymnast after a routine – we’ve all wondered about what can sometimes appear to be an arbitrary system of judging something. But we have to remember that this kind of interpretive scoring of a performance is not what is happening in the world of artificial intelligence.

When we engage with a CRE lender, our AI technology learns their own specific ideal lending profile. Over 3,000 data points are summarized into a “score” for them. The real value is in the automated collection and analysis, of course, and the score itself that is generated at the end is really just an interpretation. A computer can keep track of 3,000 data points and see the result in the data, but a person needs it to be streamlined and summarized. The artificial intelligence doesn’t start out with any idea of what is a good or bad deal, it simply compares all of the data to the lending profile that was preset by the user. By learning what kinds of deals you like, in other words, it can tell you how good a new deal should look to you. Our different customers might score the same transaction in wildly different ways, and perhaps they should.  

“You may think you know all of the factors that determine whether something fits your ideal profile, but sometimes the AI gets to know them even better than you do.”

Often times, our customers want the score explained to them. This makes sense – if your child came home with a C on their report card, you might ask them why it is a C rather than a B (or a D). The score itself doesn’t tell the whole story. This is one reason we use the type of machine learning that we do, so that we can go back and explain every data point that led to a conclusion. In other words, we can tell you exactly why your child got a C instead of an A. Some deep learning models (neural networks) don’t really allow for that kind of going back and parsing out exactly where each piece of data contributing to the whole originated. This would obviously not work in regulated environments such as banks where an audit trail is always necessary.  

Finally, there’s another advantage to teaching AI what you like and then letting it score opportunities for you. You may think you know all of the factors that determine whether something fits your ideal profile, but sometimes the AI gets to know them even better than you do. Data doesn’t lie, and I have definitely seen examples of an AI system being able to “teach” users that there may be certain patterns that they are looking for (or avoiding) that they weren’t even aware of. For example, AI might look at a large number of previous loans and be able to find patterns in the data that led to late payments or even loan default, that you might have not been aware of. Now you can simply use this insight to update your lending profile. It can be disconcerting to learn that the way you’ve been articulating your ideal profile might not be as accurate as you thought. But those of us who are open to learning something new can benefit from this greatly.

  • 1
  • 2

1010 S Coast Highway 101, Suite 105
Encinitas, CA 92024

888-521-2479
info@blooma.ai

We're SOC 2 (Type 1 & 2) Compliant!

Interested in a demo or have more questions?

Contact us

© 2021 Blooma, Inc.