Hingham (HIFS): Good Yield Curve Now – But, Always Be Thinking About the Risk of the Bad Yield Curve Years to Come
I’m writing again about Hingham (HIFS), because someone asked me this question:
I was wondering how long it would take Geoff to talk about Hingham savings, seems to tick all his boxes – very low cost of funding on the operations side and a capital conscious manager with Buffett fetish. What more could you want ?
It does have a very low operating cost. But, it doesn’t have a very low interest cost. It does now that rates have been cut to nothing. But, it probably got down to about a 2% net interest margin just before the end of the housing bubble about 13-14 years ago. It had a very rough 2006-2007.
So, the bank is set up very differently than most banks I would be interested in that I’d consider very safe. Hingham is running some serious risks by being 100% real estate focused. This is because you end up with almost no “self-funding” of your lending, because your borrowers are going to be small to mid-sized (and maybe a couple big sized) relationships where they just want a lot of money all the time to invest in more and more real estate. They aren’t going to deposit a lot of money with you. Compare this to something like the C&I side of Frost where it is going to be about 100% funded by an equal amount of deposits and borrowing coming from your customers on the commercial and industrial side. So, something like Hingham is going to need to use CDs and FHLB borrowing to operate. It is going to be very, very sensitive to cost of liabilities on the very short-term. What it is basically doing is borrowing short wholesale and then lending long against real estate. That’s very good right now. It’s just gotten to be a lot better business, because the cost of its liabilities has dropped to close to 0% with Fed Fund Rate cuts in the last year. But, that is going to be a temporary situation.
So, what more would I want to know?
Does the bank understand how it is running a unique business model that has unique risks in terms of liquidity. Like, do they understand that they can have almost no credit losses and the lowest operating expenses in the banking industry and STILL face some risks? If they said to me: “Oh. We don’t really worry about that liquidity stuff. There’s always been funds available to borrow. Etc.” Then, I might worry.
I kind of follow the Buffett approach that Alice Schroeder has talked about where I START with asking “what’s the catastrophic risk here?” and then go from there. So, if I think there’s a meaningful risk of catastrophe – then, I might pass on an idea right away that looks like it might have good risk/return odds in most environments, but could go broke in unusual circumstances.
What are the unusual circumstances that are a risk here?
Inverted yield curve.
HIFS would have a very hard time dealing with an inverted yield curve that lasted for a long time. Now, usually – at least, historically – inverted yield curves have not lasted very long. But, HIFS has gone more and more in a direction that sets it apart from other banks.
So, if you look at their most recent quarter – I think they had about 8% of their balance sheet in what I’d call “cash” and then about 2% in liquid equity securities. So, you have about 10% of the bank’s balance sheet that is kept liquid. And then you have about 90% in a mix of residential and commercial real estate loans. These aren’t liquid. They are long-term. The bank has basically eliminated the presence of any sort of bond portfolio at all. Most banks have a meaningful bond portfolio. This bank is primarily a lender – it doesn’t see itself as something that should have a bond portfolio.
The simple way of looking at that is asking is it good or bad?
The answer to that simple question would be: “it’s good”.
A liquid equity portfolio of $50 million in common stocks is a better long-term asset (it will compound at a higher rate) than a government / agency / municipal / etc. bond portfolio of $50 million. I mean, you could get capital gains / dividends of $2.5 million to $5 million a year (like $1 to $2.50 a share in annual “comprehensive earnings”) coming from an equity portfolio and much lower from a bond portfolio. So, over time, a bigger weighting to equities instead of securities is better here.
And then, of course, the loan portfolio will have a higher yield (net of losses) than a bond portfolio would have. However, the loan portfolio has to be seen as illiquid. So, this is a leveraged and illiquid entity.
It also, because it’s 100% a real estate lending business, always going to be borrowing short and lending long.
So, I look at the 25+ years this bank has been operating this way and think: has this been a good time generally to be borrowing short and lending long?
And the answer is, with the exception of the housing boom – yes, it is.
So, does that mean the past record isn’t a good indicator of the future?
Well, for the immediate future: it’s fine. The immediate future at HIFS will be excellent. But, what if short-term rates rise a lot vs. long-term rates and stay that way for a long time.
My feeling on that after analyzing the bank – by “analyzing” I just mean I read everything on their IR page (so 25+ annual reports, presentations to shareholders, etc.) that the bank is in a better position now than it was in 2006-2007.
ROE probably dropped to like 8% or so in 2006-2007.
That’s not a total deal breaker for a bank.
But, if you’re talking about paying 1.7 times book for something that will sometimes return 8% on equity. That is a deal breaker if there are a lot of those years. If we divide 8% by 1.7 we get about 5%. Now, do I think equities will do much better than 5% in terms of indexes and such over the next 10 years?
Maybe not. But, indexes might do somewhat better.
Okay.
But, what if there’s a period where short-term rates are high and long-term rates are low relative to short-term rates (not inverted for 10 years, but a “flat” curve).
Well, indexes and stocks generally probably won’t do too well. And they might not do better than HIFS.
And then if you have a steeper curve – HIFS will do a lot better than index funds even if you buy at 1.7 times book value.
When looking for a stock, I generally want two things:
1) A really low risk of “catastrophe”
2) A reasonable chance (always hard to define this in terms of probabilities) of getting a 10x return in 15 years (that’s a 17% CAGR for 15 years)
We could simplify #2 for rounding purposes to basically asking:
– Can this stock (under some reasonable scenario) compound at about 15% a year for about 15 years?
It doesn’t all have to be from compounding of book value. Some could be from dividends. And some could be from multiple expansion.
Let’s take a look at Hingham.
How close can it get to returning 17% a year over the next 15 years.
Dividends can add: 1% a year (based on recent dividend yield)
Multiple expansion (higher P/E at end of the 15 years) can add: 3-4% a year (P/E goes from about 9x forward now to about 15x forward at the end of the period)
That gets me to just a 4-5% return.
Now, we need compounding of about 10% a year.
BV has compounded here at 15% a year for a long time. So, sounds okay.
But, the balance sheet has expanded to a pretty big size now. So, it could be harder to grow.
Can they keep dividend payout ratio at like 30-45% for the next 15 years?
Given their historical and current ROE and dividend payout and growth…the answer would be basically yes.
If ROE is about 15-18% and dividend payout is about 30-45%, growth can be about 10% a year in BV (that 10% a year is what’s left over if you do 1-dividend payout ratio).
Then, there’s the question of leverage.
Can the bank’s leverage ratio stay about even or rise over 15 years?
Seems as likely to rise as fall. Seems like a normal level.
So, answer is yes.
So, I’d say you can look at the past record and try to learn about the bank and talk with management – and, yes, you get an answer where 15% a year return in the stock (but not the actual business itself) is in line with expectations over 15 years.
It hits my return threshold.
But, does it clear the risk threshold?
We can’t predict the future. And I don’t start by saying: “What do I think the yield curve will look like for the next 10 years?”.
Instead, we have to just weigh the possibilities of whether a damaging yield curve is likely for a long time. And whether trouble accessing capital is an issue.
So, then, the questions generally end up being focused on things like:
– Can you cut costs further over time (if net interest margin got to 2% before, but you can take another 0.4% out of your operating expenses that can make it like net interest margin is equivalent of 2.4% now vs. then).
– Is the bank in any way reckless about liquidity risk
They aren’t reckless about credit risk. I know that from talking to them. I know that from the past record more so than anything else.
But, they’ve had success being completely in real estate lending for 25+ years and getting a less and less liquid balance sheet over time. So, it’s easy for inertia to set in and believe in your own success and end up thinking: “We always run this kind of liquidity risk – it’s never a problem.”
So, the answer is what more can I know?
I can know: how serious are they about managing their liquidity risk. How vigilant will they be even after many, many years when they run this risk and face no negative consequences from it. And I can also ask about things like deposits. Can you grow deposits over time in your “specialized deposit group” and things like that to do anything to lessen these risks.
What you want to hear is that they are vigilant. They know it’s a risk they are running that some banks aren’t. You want to hear they are serious about growing the specialized deposit group. And you want to know they are serious about continuing to take costs out of the business to make the really, really flat yield curve years (and especially the periods where the curve is inverted) survivable.
The risk that I see with Hingham is just that you could encounter a period with the yield curve that is so bad for borrowing short and lending long and more importantly – lasts so long – that it permanently alters their compounding rate. If every couple decades you hit two years where ROE goes from 15% to 7.5% for two years – this is something that can be overcome. But, you go much beyond that (if these moments occur more frequently, if they last longer, etc.) you have a problem for long-term compounding.
Take costs out is an important part of that. If the next really tough period is 10 years from now and they manage to lower operating expenses from 0.8% of assets today to 0.6% of assets in 10 years – then your ROE could go from 7-8% in those bad years to 9-10%. I can easily live with an ROE close to 10% a year for several years in a row.
But, if it goes the opposite way – where they aren’t vigilant about costs and they allow operating expenses as a % of assets to creep up from 0.8% to 1.1% or something – now, you are (other things equal) at a sub 5% ROE. If you have to live through a period like that which last 3 years or so on average – it makes a big difference whether the results are 15% ROE 15% ROE, then 5% and 5% and 5% and back to 15% vs. 15% ROE, 15% ROE, then 10% and 10% and 10%.
Connected to this.
What about dividends and share buybacks?
What I want to hear is that we don’t want to keep raising our “regular dividend payout ratio” (HIFS does a special dividend too) and we want to add stock buybacks to our arsenal.
This is really important in bad yield curve years. Because what would happen if you had a 30-50% regular dividend payout all the time you never were willing to get away from is that you’d spike to like 65-100% dividend payout in the bad years.
If you’re willing to really keep the dividend payout low and to consider stock buybacks, then you can buyback your stock in the bad years.
Aside from one-time events like COVID earlier this year – my guess is that the forward return expectation on HIFS stock will exceed the returns in the business only in these inverted or flat type yield curve situations.
So, it’s possible that you can do your best capital allocation in bad yield curve years in the form of buying your own stock. This is because the decision of whether or not to buy the stock should be based on solving for this inequality:
Long-Term ROE Average / (P/B) > Expected Return on Index?
For today, assume:
15% ROE
1.7x P/B
“x” expected return on index
15%/1.7 = 9%
It’s not as simple as I laid out. But, we can simplify to 9% in the sense that if 9% is greater than return in the index and we buy – we haven’t made a mistake. The mistake we could make due to our oversimplification is only NOT to buy HIFS when we should not the other way around.
So, for today – the inequality simplifies to 9% > S&P 500 forward return?
Yes or no?
My gut says yes. But, you can answer for yourself. I don’t really look behind 15 years when thinking about a stock investment. So, when I say “forward return” on S&P 500 – I mean from 2020-2035 or so. Will it be 9% or higher (and will HIF’s geometric ROE – not arithmetic – be 15% or higher).
How does this influence the buyback math?
Say you expect an 8% return in the S&P 500 normally. Your stock has a normal ROE of 16%. Then, your stock might be “fairly valued” around 2x P/B.
So, buybacks at small discounts to 2x P/B aren’t a big deal. Do them. Don’t do them. Who cares? I mean buying back 2% of your stock at 1.8 times P/B isn’t going to move the needle for your shareholders. But, having a chance to buyback 10% of your shares at 1x P/B really does move the needle.
Will you ever get such a chance?
Right now, it seems like the answer would always be no with a stock as well thought of (in terms of P/B multiple premium over other banks). But, let’s consider when this opportunity could present itself…
…What if the market responds more to the CURRENT return on equity you have instead of your long-term average. You ROE drops from 16% to 8% and the stock also drops by almost 50%. This sounds crazy. But, it means the market would be keeping a constant P/E on the bank. Personally, I think using a P/E from a non-representative year of ROE is crazy. But, I can imagine it happening. Sometimes, even stocks see their P/Es contract in an especially low ROE year – which means the “normalized P/E” has plummeted tremendously.
Well, now, your stock is very undervalued. Because the yield curve will change again and you’ll again be making more like a 16% ROE than an 8% ROE.
You shouldn’t be paying dividends then. You should be buying back your stock.
So, the “more” I want to learn is:
– How seriously do they take liquidity risk?
– How low are they willing to keep dividend “commitments” to shareholders?
– How willing are they to do big share buybacks instead of dividends?
– How focused on growing the specialized deposit group are they?
– How focused on continuing to lower operating expenses even if they are the best of class in this area already?
– Finally: how good are they are thinking about catastrophic risks when year-after-year they are running risks that don’t show up in results at all
The hardest one is the last one. If a business has no ill-effects from some risk it is taking for like 10+ straight years – that risk becomes invisible to most managers. It takes very, very special managers to remember they are running invisible risks that no one is thinking about or has thought about for many years.
So, vigilance and honesty about something that can’t be seen in the actual past experience record is really the #1 thing.
I can see how a bank like HIFS could be run well and have a great compounding record for a long time and then be undone by a risk “no one saw coming” – even though everyone should know it will come someday. It’s a super infrequent risk. But, it has a high magnitude of loss in that moment when it does happen. So, you don’t want management at someplace like HIFS to just wave off the idea of longer-term inverted yield curve than we’ve ever seen before. You don’t run the business every day thinking about that. And you shouldn’t. But, it’s very easy to just assume you will never face the one catastrophe that could undo everything you’ve done over a lot of years.
So, I guess judging vigilance for a seemingly “invisible” risk is the more I want to know.