Artificial intelligence and data centers have dominated the conversation about where capital is flowing, but for many middle market borrowers, the AI boom is a sideshow. Manufacturers, staffing firms, engineering services companies and packaging businesses still need lines, capex facilities and a lender that understands their business.
Mike Semanco, president and COO of Business Finance at Mitsubishi HC Capital America, has spent his career lending through cycles, including the Great Recession, the 2023 bank liquidity scare and the rush into electric vehicles. Semanco runs the company’s asset-based lending division, which serves a broad group of generalist borrowers, some directly tied to AI and many simply using it as a tool.
In this conversation with Rita Garwood, editor in chief of ABF Journal and Monitor, Semanco explains why collateral still comes first, which metrics matter most now, how his team decides whether a pivot is prudent or overextended, and what borrowers without an AI story should do. His advice is to stick to your knitting.
Read or watch the full interview below or listen on Spotify.
Rita Garwood: What are you seeing in the market right now? Where is capital flowing, and how much of it is tied to AI and data centers versus the rest of the market?
Mike Semanco: It’s an interesting dynamic. Every article or publication seems to be based on some type of AI, whether directly, through a data center or construction, or through some ancillary product related to it. That is positive. We have a structured finance team within Mitsubishi that focuses on data center and project financing, which has been very beneficial for them over the last couple of years and is really driving their business.
On the asset-based lending side, which is the division I run, we’re seeing a variety of companies. Our focus is more on generalists, so we’re seeing both companies directly related to AI and companies using AI as a tool, whether it’s a staffing business, engineering services or manufacturing. They’re not fully engaged in AI, but they have that AI conversation in their business plan. So it’s been a mixed bag. But when we look at the non-AI companies as borrowers, they have just as much potential as the AI-related companies.
Garwood: AI and data center deals are getting a ton of attention right now. Are you seeing that attention change how capital is priced or allocated for other types of borrowers?
Semanco: Not from our perspective. Our objective is to look at all borrowers that come to us and put together a solution to help them grow their business. We’re in business to make them more successful through our type of financing. We haven’t changed our capital allocation or pricing for companies that aren’t AI-related. We look at the underlying fundamentals of the borrower, understand their business plan, and understand their needs and objectives. Whether they’re AI-related, on the fringes or using it as a tool, we look at the borrower as a standalone business we can help be more successful.
Garwood: How would you describe the state of the middle market right now, outside of the AI conversation?
Semanco: I use the term “mixed bag,” and that’s true across industries. We’ve seen manufacturing and packaging companies that are more equipment-intensive ask for increased lines or larger capex facilities because they’re seeing growth. At the same time, they’re faced with higher costs from tariffs and energy prices, including oil and gas. So there’s upside, but there’s also uncertainty about where things are going.
We see the same thing with receivable-based businesses, such as staffing and engineering services companies. Their customers want skilled talent in addition to leveraging AI. Those businesses are growing, and they’re looking at how to do more with the true talent they have. It’s the same in industries like healthcare, where AI and technology-based solutions are much more efficient and help the consumer. Those businesses are growing, but there’s still some reluctance because of geopolitical movements and uncertainty in the market. There’s cautious optimism, and what we’re seeing makes sense.
Garwood: Let’s talk about underwriting. How does your company underwrite a business with strong collateral and cash flow but no AI narrative? What are you looking at first?
Semanco: For our ABL clients, it’s all about collateral first. We focus on receivables, inventory and equipment. Some clients are heavier on receivables and inventory, others on hard assets. Our discipline moves us to collateral first, cash flow second. As we move through the market and look at various stages of our clients, there may be situations where, even though we’re collateral-focused first, we need to stretch a little. Then we focus on cash flow and free cash flow, and on whether the company can sustain a downturn or the loss of a contract.
The collateral may certainly be there, but the cash flow is potentially more important if a contract is moving around, or may have to grow or shrink. So it’s always collateral first. We look at those values, but then we rely on cash flow and free cash flow to let us stretch, if we have to, for a borrower that’s growing.
Garwood: Which performance metrics matter most to you right now? Have any shifted up or down in importance over the past year or two?
Semanco: Beyond collateral values, free cash flow and operating income are the two metrics we focus on. Both show what a business can do from a resiliency standpoint. If there’s a downturn or something happening in their industry, can they absorb some of those shocks?
In the last six to 12 months, EBITDA add-backs, or EBITDA adjustments, have been a key focal point, not only for us but for other private credit lenders as well. What is the company adding back to arrive at a true operating history? We still look at EBITDA, and we look at it from the perspective of where the company is going. But as a pure metric, free cash flow and operating income give us a direction on whether the company is going to face a challenge.
Garwood: Going back to AI for a moment, are you seeing borrowers feel pressured to add an AI narrative to their pitch if they’re trying to attract capital? How do you respond if they do?
Semanco: We haven’t seen that. In some of the syndication transactions we’re in, the lender pitches, even from large companies, do mention AI and how they’re using it internally. But I don’t think companies are being pressured. I think they choose to include it to show how they’re thinking about the business going forward. In some businesses it’s more important to have that AI component in the lender pitch, the projections or the borrower deck. In other industries it’s less important. The clients who add it see it as a beneficial tool.
I can’t think of companies that are totally pivoting from what they do day in and day out to an AI business. We’re seeing clients use it as a tool, from an efficiency standpoint or to better serve their customers. That’s where we’ve seen AI mentioned in their lender decks.
Garwood: You mentioned collateral is the first thing you look at. How do you evaluate collateral in a market where asset values in some sectors are being driven by the hype around AI?
Semanco: It’s interesting. I think it started outside of AI, when the tariffs hit earlier in the year and drove up inventory values. We looked at where the inventory value was for a company, what its raw materials were, and took a deeper dive to understand the true cost. The same goes for AI-related capital assets and that industry. We’re fortunate to work with a lot of appraisal companies that live that business day in and day out, and we lean on their expertise. We also have an asset management group within Mitsubishi that we can tap to understand where equipment or inventory values sit today.
Leveraging professional partners who do this often, with a skill set and knowledge base beyond ours, means we may come into a transaction with a lower loan-to-value on a particular asset class but give the borrower more credit in other areas. Maybe we stretch if there’s free cash flow or a new contract, or we offer a higher advance rate on receivables. We look at the full picture of the business and try to maximize its value, but within a credit box we feel comfortable with based on market conditions.
Garwood: When we were planning this podcast, you mentioned that Mitsubishi HC Capital America kept lending to manufacturers when banks pulled back during the Great Recession. What made that possible, and what did you see on the other side?
Semanco: It was certainly an interesting dynamic, being in the Midwest with heavy connections to the auto industry. There were challenges that ran through that, including multiple bankruptcies. We took the time to really understand the client base we were working with, where they were in the supply chain, how their customers saw them, and which ones had the wherewithal and the capital base to survive going forward. A lot of it had to do with customer relationships.
I remember we helped a manufacturing company in auto that no bank would touch. A lot of it came down to how supportive their customer was, what platform they were on, and how they could deliver given the customer support they were getting during that period of uncertainty. Having those connections and being involved in those conversations allowed us to get comfortable. It’s a deeper approach to the client, getting involved in their business from the owner’s perspective and understanding where they’re going. That allowed us to continue to lend in that market. Not every opportunity was positive for us, but getting closer to the client and taking a consultative approach really helped us through those difficult times.
Garwood: What did that period teach you about the difference between a temporary disruption and a structural one?
Semanco: It was interesting to lean on that process from the 2008-to-2010 time frame and compare it to the bank liquidity crisis in 2023, because a lot of similarities came up. I remember discussing internally, with Silicon Valley Bank and others in the news, whether the ripple effect could spread across banks as a whole. Our analysis was that it wasn’t a structural issue. It was limited to certain institutions. We didn’t see a bank liquidity crisis or a run on banks across the globe, so we took a different approach. We had good relationships with the banks we were doing business with from a syndication standpoint. It was a point in time, not a systemic issue that touched everything, as the global recession did.
Garwood: Let’s talk about EVs. Before AI, they were the hype thing everyone was talking about. Some manufacturers retooled and invested heavily to chase those opportunities, and those haven’t necessarily materialized as initially expected. What did that look like from the lender’s seat?
Semanco: When you see a new industry developing and morphing from ICE vehicles to EVs, there was a lot of emphasis and investment from some larger suppliers and the OEMs. From the lender side, it was about getting back to the discipline and fundamentals of what we do day in and day out in ABL: looking at where clients are today and then at projections of where they’re going.
Some of our hesitation was that these companies were so good at what they had done over the last 20 or 30 years. Some were dipping a toe in the water for EVs, and it felt like a calculated risk. Others were pivoting 100 percent. That’s where we saw hesitation from a lending perspective, when they were going all in because EVs were the next industry they were going to try to capture. The outcome wasn’t detrimental across the board, but some companies struggled quite a bit because the pivot was so dramatic.
It gets back to what we talked about earlier. If you have a company that isn’t familiar with AI and it’s all of a sudden all AI, is that beneficial? Are they going away from the true business model they created and were successful at, and taking on something new that could carry a lot of risk? In my history in lending, whenever I’ve dabbled in a credit I know very little about because it sounds interesting, that’s usually where the credit losses come from. I tell our people to have the knowledge base and not to dabble in industries or asset classes we know nothing about, because usually that’s where we pay the price. We saw the same thing with the conversion from ICE to EV. Some did it strategically and calculated, and others jumped all in and didn’t fare so well.
Garwood: What were the warning signs, if any, that the pivot to EV was overextended? Did you decline any of those deals, and why?
Semanco: We did. The signs were overly optimistic projections, most likely based on what the client was hearing from their customer. There was a lot of projection about how fast production would be built and how quickly the market would absorb it, while at the same time consumers weren’t going all in. Projections that seemed out of step with what the underlying buyer was focused on were a red flag for us. That included companies trying to piggyback on others that were farther along, or that were in the early stages of work on battery-powered trucks and other vehicles, and trying to match or exceed what an existing business had done for years. The unrealistic expectations threw us off when we looked at business plans for EV clients.
The second thing was, do they get paid? We can lend money, but ultimately we have to get it back. Because we’re collateral lenders, we looked at the underlying collateral and whether it was turning in the time frame we wanted. Some of that was being stretched out, and it created uncertainty from a credit box standpoint. So we did pass in some situations where it just didn’t make sense for us. A lot of it had to do with expectations, and the projections seemed very inflated.
Garwood: How did the companies that stayed the course fare compared with those that pivoted harder?
Semanco: It had a lot to do with capital. The companies that dipped a toe in the water and turned a segment of their business into producing product for EVs, whether cars, trucks or work trucks, fared a lot better because their capital wasn’t solely invested in another platform when the pivot ultimately happened. It was much more calculated, and the calculated risk these companies took ultimately played out.
I think it’s based on their history. They’ve been through downturns and situations like the Great Recession that really impacted them dramatically, when a lot of car companies, or the economy in general, softened. They learned from that. They know how the story ends. So the calculated model, and having gone through different cycles, certainly helped those companies versus newer ones, maybe startups, that didn’t have that experience.
Garwood: You’ve said AI is creating both disruption and opportunity, much like the recession did for manufacturing. Where do you see the parallels, and where do they break down?
Semanco: A lot of it comes back to the calculated risk I mentioned for the recession. Companies in manufacturing, services, packaging or another industry that isn’t directly technology-based: are they going into AI as an end-all, be-all? Are they eliminating staff and leveraging AI to do all the work staff used to do? To me, that’s an all-in scenario.
Then there are companies that use it as a tool to enhance their business. We have some service-based companies using AI to serve their clients better and win more business. Very similar to manufacturing in the Great Recession, those are the ones that will come out ahead. There will be technology companies that pivot and do AI 100 percent. I’m not knowledgeable enough to say whether that will work. I can only tell you that, based on our client base and the middle market companies we work with, they’re using it as a tool rather than making a complete pivot to becoming an AI company.
Garwood: You’ve suggested that software and tech companies that don’t go all in on AI might be better off, like the EV situation. Can you explain that thinking? Is there a version of AI investment you would consider prudent?
Semanco: I have a limited knowledge base of AI from an investment standpoint. But for me, running a business means sticking to your knitting: understand what you’re good at and focus on the customer, both existing and new, and on the product offering that gives them value. If you’re in a non-AI-heavy technology business, can you leverage AI, or technology in general, to make your company more efficient and more attractive to new customers? I think that’s where the AI model helps companies, versus a complete pivot.
I think the software companies that use AI as ancillary to what they’re doing, as a tool that enhances it, will do better than those that completely pivot. There’s going to be a bifurcation of big companies that are AI-focused and smaller startups. Some may have the right model for a larger company to buy, because we’ve seen that before, but not all the startups creating AI are going to make it. At the end of the day, it’s a calculation, and part of it is that you don’t know what you don’t know because it’s so new. Calculated risk, at least in my mind, and that’s the credit person talking, seems to have a better outcome.
Garwood: How would you distinguish a company making smart, measured AI investments from one that’s just chasing the trend?
Semanco: I’m probably not the AI expert to answer that. But the companies we see that are calculated, using AI as a tool to help their business rather than pivoting entirely, are going to be better off, whether through the business they can do directly or because they could become an acquisition target for a larger business that can’t get into AI itself. We have clients using AI as a major tool to help their customers, because those customers haven’t been able to figure it out on their own. To me, that’s a value.
A larger company that uses a smaller business with the mobility and flexibility to leverage AI may end up with an acquisition target on its hands. It’s easier to buy something than to create it. The larger business avoids making a large investment in something it’s not familiar with and partners instead with smaller companies that have flexibility and an entrepreneurial mindset. That’s good for the overall ecosystem.
Garwood: What risks do you see if the AI investment cycle cools down or concentrates?
Semanco: I don’t know the complete parallels, but I think it will be similar to the EV space, where you’ll see a lot of consolidation. The larger businesses that have the capital can weather the storm and keep making those investments, versus smaller companies trying to get into AI full time. The competition is going to be dramatic, and there may be acquisitions of smaller companies by larger ones as add-ons, which will be interesting.
There are going to be a lot of unknowns. It seems like we read every day about what’s happening, its longevity and who’s driving it, not only the AI companies but the chip companies and cyber companies. There are a lot of dynamics in that technology space that I’m not smart enough to have all the answers for. It’s going to be interesting to sit in the stands and watch the show play out.
Garwood: What should borrowers without an AI story do now to position themselves for capital? Does it even matter if they don’t have one?
Semanco: We’re seeing a lot of companies that aren’t directly related to AI, and I don’t think they need an AI story. They have a story about their business and what they’re doing to help their customers be more successful. If they can plug AI into it and make themselves more successful, so they can make their clients’ business models more successful or attract new clients, that’s how they can position themselves. But we’ve seen a lot of companies recently that don’t have any type of AI story. It’s behind the scenes, not front and center. They’re doing what they do best, and they’ve created a sustainable business model.
We’re still lending to those businesses because they’re good operators with a good business model and good customers. Again, our objective is to make them more successful with our financing. That’s what we’re in business for.
Garwood: What should lenders and other capital providers be doing differently, or avoiding, as this cycle plays out?
Semanco: The beneficial part of asset-based lending, and what we do day in and day out, is the discipline: the tried-and-true discipline of understanding the cash flow of the business, being a consultative partner to your borrower and seeing them through the ups and downs. I think lenders should have that mentality of being more consultative. There’s always going to be a lender willing to take on a little more risk than somebody else, so there will always be money available to borrowers. But the lenders with long-term staying power have that discipline, the fundamentals of ABL: collateral first, cash flow second, and looking at character and the operator. Those disciplines make our business, and other lenders’ businesses, a little more stable as the economy moves or as companies shift.
Garwood: Last question. Looking out over the next 12 to 24 months, what would change your view of where capital should be flowing?
Semanco: Having a little more certainty in the market. It’s bifurcated: a lot of larger companies have positive earnings, but some of the smaller companies are dealing with uncertainty, whether it’s rising costs or labor issues and being able to hire good talent. During the Great Recession, when the automotive companies were more sure about where they were going, the supply chain was more sure, too.
So I’d like to see more stability over the next 12 to 24 months: maybe geopolitical movements calming down and not being so much of a distraction, and AI finding its footing with whatever guardrails the smart AI creators put in place. It’s about certainty, and not having the small to midsized business owner face so many things that keep them up at night. That’s what I’d like to see, a little more stability.
Garwood: We’d all like to see that. Mike, thank you so much for joining me on the podcast today and talking about this topic. I really appreciated this time with you.
Semanco: Thank you, Rita. I appreciate you having me on, and I look forward to future conversations.
