Recently, I wrote an article for SNL Financial in which I discussed the competitive, operational and economic factors that have worked to depress bank M&A activity since the beginning of 2008. Even though activity is depressed, some deals have been getting done. While you might expect the deals to be priced and structured conservatively given the ongoing economic malaise, you'd be mistaken. Some mediocre franchises have sold for what I consider full prices. That's a bad sign for acquirors' shareholders, both those on the "winning" side of recently announced deals and those who hope to buy at reasonable prices in the future.
On Friday, Short Hills, NJ-based thrift Investors Bancorp, Inc. MHC (NASDAQ:ISBC) announced its acquisition of Astoria, NY based Marathon Banking Corp. for $135 million in cash. The price equates to a multiple of 1.51x Marathon's tangible book value per share. Is that high or low? That depends on two related things: capital levels and return on tangible common equity (ROTCE). If Marathon has excess capital, it seems logical that a buyer shouldn't pay a premium for that excess. Second, the higher the return on equity, the more valuable the institution, and therefore the greater the price/tangible book value ((P/TBV)) multiple a rational buyer should be willing to pay.
Marathon's tangible common equity/tangible assets ((TCE/TA)) ratio at Q1 2012 was 10.14%. So Marathon is overcapitalized. If we assume a "normal" TCE/TA ratio of 7%, that would imply Marathon has $29.7 million of excess capital. If you remove that excess capital from Marathon and also from the deal value, and recalculate P/TBV, you get 1.77x.
1.77x feels rich, but whether it actually is rich depends on ROTCE. Since 2010, Marathon's ROTCE has been a very low 7-8%, and not because of a high provisions/average assets ratio, which is the reason many bank ROTCEs are currently depressed. And it's not because of a depressed net interest margin or high overhead relative to assets either. It's largely because of mediocre fee income. A 1.00% ROA bank (about average) with a 7% TCE/TA ratio delivers a 14.3% ROTCE. If Marathon had an average fee income/average assets ratio (1.24%) rather than what it actually has (0.27%), its ROA would be 1.35% instead of 0.72%. Remove the excess capital and Marathon's Q1 2012 ROTCE increases from 7.9% to 11.5%. Better, but still far from good.
Of course, ISBC plans to realize synergies from the Marathon acquisition. That's why ISBC, or any acquiror, is willing to pay an acquisition premium. In the investor presentation on the deal, ISBC stated that it expected to realize synergies equating to 30% of Marathon's non-interest expense base. That equates to $4.5 of fully phased-in after-tax synergies. Those synergies would take Marathon's Q1 2012 non-interest expense/average assets ((OH/AA)) ratio from 2.55% down to 1.79%. 2.55% is already lean - 1.79% is emaciated. In Q1 2012, ISBC ran at a 1.65% OH/AA ratio. But here's the problem; ISBC is a thrift. Marathon is not a thrift. Thrifts run at lower OH/AA ratios than banks. The synergy estimates in this deal are aggressive.
Here's the other problem. Marathon's Q1 2012 stand-alone earnings annualized are $6.5 million. $4.5 million of after-tax synergies will increase Marathon's stand-alone earnings by 68%, to $10.9 million, and its ROTCE from 7.9% to 12.9%, if unadjusted for the removal of excess capital, and from 11.5% to 18.7% if adjusted. Even if ISBC thinks it can realize this level of synergies, should it be paying Marathon for all of them? Why bother doing the deal?
How can we assess whether ISBC is paying for all of the estimated synergies?
The 30-year Treasury bond yielded 2.67% at the close of business on Monday. All other things being equal, the lower this rate, the lower the appropriate discount rate for equities. A 5% equity risk premium would imply a 7.67% equity discount rate for a company with a beta of 1. Of course, the "fair value" estimates derived from such a low discount rate are highly vulnerable to interest rate increases. This discount rate is too low. In my opinion, bank stock investors are currently using a discount rate in the neighborhood of 10%. And that implies Marathon deserves a 1.87x multiple on its slimmed-down tangible common equity of $60 million, if it can hit those aggressive synergy targets. That gives a valuation of $111 million. Add back the $30 million of excess equity capital back and you get $141 million. That's a $6 million excess above the $135 million purchase price (a 4% increase), and it's the reward ISBC shareholders get if and only if those synergy targets are hit. Spread over ISBC's 112 million shares outstanding, it translates into a $0.05 per share benefit.
How can this be, if ISBC's investor presentation touts EPS accretion of 8-9% beyond 2014? For the simple reason that ISBC is paying cash for Marathon. Cash is a superior acquisition currency, especially in today's low-yield environment. ISBC's Q1 2012 pre-tax yield on interest-earning assets was 4.55%, 2.96% after-tax. If we assume this is ISBC's cost of cash, ISBC could pay about $369 million in cash for Marathon's $10.9 million in pro forma earnings without realizing EPS dilution. That's obviously a ridiculously high price. In contrast, if ISBC were to use its stock to buy Marathon, ISBC could pay only its own P/E for Marathon, which would yield a break-even price of about $182 million. There's a difference between what an acquiror is capable of paying and what it makes sense to pay. EPS accretion measures the former. It sometimes does an acceptable job of measuring the latter, unless the acquiror is using cash consideration, as the arithmetic above clearly demonstrates.
Let's return to that 10% discount rate for a minute. Because ISBC's investor presentation claims that the Marathon deal would deliver an internal rate of return (NYSE:IRR) of 18%. That's a nosebleed IRR in today's low rate environment. It's also highly misleading, if not useless. It has become fashionable in some recent bank M&A deals to tout high IRRs. Hedge funds and private equity firms report IRRs when they sell, not when they buy. And IRRs have meaning for the investments these firms make because the expected holding period is finite and usually short. ISBC doesn't plan on selling Marathon; does this mean that the 18% IRR is perpetual? Of course not. I can get close to the 18% IRR figure, if I assume a finite holding period and also that the exit multiple on pro forma earnings is equivalent to the multiple on pre-synergy earnings that ISBC is paying for Marathon now, which is completely unrealistic.
You don't see these calculations discussed in merger proxies, I'm guessing because such a discussion would highlight how flawed they are. Investors are meant to take them at face value.
So can you tell I think ISBC overpaid? By how much? I'd be happy giving Marathon shareholders credit for 15% synergies. This is low relative to synergies assumed in other recently announced deals, but Marathon already runs lean. ISBC shareholders would then capture any additional upside. I don't think acquirors should make aggressive synergy assumptions and then pay all the value of these synergies out to the target shareholders. Doing this, and then touting the deal's "strategic value", isn't visionary, it's destructive. 15% synergies equates to $3.4 million of pre-tax earnings, or $2.2 million after tax. Those synergies would increase Marathon's earnings by 34%, and its "slimmed down" ROTCE would increase to 15%. And that would justify a purchase price of $120 million, $22 million of which would represent the after-tax value of the synergies.
So, I calculate that ISBC overpaid to the tune of $15 million, or 12.5%. If your response is "who cares?" perhaps because ISBC has over $11 billion in total assets, there's the problem. Would ISBC shareholders be angry if an ISBC employee stole $1 million from the bank? Of course. Then why shouldn't they be more mad if ISBC's management team squandered multiples of this? The smartest acquirors think about this the right way. They buy only when price is less than intrinsic value, even though they might end up being less acquisitive as a result.
Discipline is a rarer management attribute than "vision." I'm curious which we'll see more of when bank M&A activity eventually picks up.