Cure For What Ails Banks May Be Less Securitization And Trading, Or Much More Securitization And TradingBy Matt Levine
There’s a lot of The Future Of Banking in the news today and we should talk about it but first a proposition. Where are you more likely to lose money on a mark-to-market basis: buying 5-year PIK-toggle holdco Petco bonds at 8.6%, or buying 30-year UPS bonds at 3.625%? I say your odds of losing on UPS are higher; if you disagree, you take Petco, and we’ll meet here in 30 years to settle up.
Here is a grab-bag of other numbers related to UPS:
- It paid its underwriters 87.5 basis points to underwrite those 30-year bonds, or about 3 basis points per year of financing.
- Its credit facility pays its banks 4.5 basis points a year on undrawn amounts; on drawn amounts it pays Libor + 10 or more basis points.1
- It has $1.775 billion of commercial paper outstanding with an average interest rate of 0.07%.
Things to think here include:
- lending money to UPS is not profitable for banks2;
- underwriting UPS’s 30-year bonds isn’t exactly a bonanza either; and
- UPS would be nuts to borrow from its banks – so it doesn’t, and borrows more cheaply in the market.3
Bank lending to high-investment-grade companies is (1) a loss leader, (2) used to attract not especially profitable business, and (3) not competitively priced. I feel like other industries do loss leaders better.
While you ponder that, also ponder this IMF working paper on banks and trading. A quick takeaway is “banks shouldn’t trade, urgh, trading bad,” and Mark Gongloff and Felix Salmon take that takeaway away, but as far as I can tell the more interesting bit is this:
In the last 10 to 15 years, financial markets have deepened substantially and traditional banking has become less profitable. The two trends had the same driver: information technology has increased the availability of hard information, expanding the universe of tradable claims and making banking more contestable. … To put it in starker terms: because of financial development, the business model where a bank combines core relationship operations with a transactional activity – be it in traditional European universal banks, in commercial banks that hold securitised products, or in investment banks – is no longer sustainable.
Trading in modern banks opens the door to risk-shifting and hence will lead to bank failures. Trading also leads to a misallocation of resources from lending; this makes banks unable to maintain relationships and leads to a reduced supply of customer-oriented services, such as SME lending. Trading then also compromises the role of banks as providers of liquidity during economic slowdowns.
Also, there’s this McKinsey study out that says a bunch of things, including that banks should cut comp to increase ROEs, which you may not have been expecting. You, and Bloomberg, of course care about comp, but that’s not all. There’s also buzzwords:4
However, the relative value of corporate lending is declining because banks no longer enjoy structurally lower funding costs than many of their large corporate clients.5 Leading banks are responding by pushing cross- and up-selling, transforming front-office processes, applying lean solutions and adopting e-solutions. …
The capital markets business is the most challenged segment, due to regulatory pressure, higher funding costs, and shrinking revenues. However, capital markets have responded faster and more radically than some other segments, in particular over the past year. …
Still, the adjustment challenge should not be underestimated. Lower revenues and higher capital needs mean the cost base must be cut radically. Further, some fundamental questions need to be answered, for example how the capital markets business will be funded going forward (a particular concern for banks without a solid deposit base).
McKinsey’s recommendation is not just “cut salaries” but also amazingly this:
Banks must review loan books, enhance risk models and improve collateral management. In addition, they must implement structural changes, for example by shifting financing off balance sheet. This is particularly true for European banks, which in 2011 had far lower ratios of securitized loans and corporate bonds to total financing volumes (19 percent) than US players (64 percent).
Although the IMF economists and McKinsey come to amusingly opposite conclusions – “less securitization” versus “much, much more securitization” they’re also sort of saying the same thing. Basically the model is:
- If you are the Bank of Bedford Falls,6 you (1) take deposits at 3%, (2) lend to upstanding people at 6%, and (3) don’t lend to miscreants at any%, and (4) thereby make enough money to pay your CEO a decent salary.
- If you are the Bank of, not Beford Falls, but oh, say, America, or Deutschland for that matter, you can’t really scale that into enough money to pay your CEO an indecent salary, because (1) weeding out local small-business miscreants is just as labor-intensive as it is in Bedford Falls and (2) the upstanding people who actually want to borrow billions of dollars can do so cheaply from the public markets.
- So how do you scale? One way is just “lend to miscreants,” and to some extent that’s the answer, but you’ve also presumably gone to finance school and know that you want risk-adjusted returns. So the real answer is “pass the miscreant risk on to others,” in the form of securitization or hedging. A further answer is to take your deposits and use them to fund adventures in the markets, which are positive-expectation if you fund more cheaply than everyone else in the markets, which you used to, though now you sort of don’t, mostly because you did this.
McKinsey basically takes this to its logical conclusion – just be all-markets, no-capital7 – while the IMF researchers prefer a return to relationship-based, soft-factors-driven lending, without the market appendages. The IMF paper’s answer is right, in the sense that banking – relationship-based financing of people and businesses without access to the public markets – is an actual thing with social value distinct from the pure market-driven risk-pooling function offered by banks but also by RMBS pools and corporate-bond ETFs. McKinsey’s answer – basically, “become securitization engines” – seems sort of sad, but also far more likely to actually happen.
Companies Feast on Cheap Money [WSJ]
The risks of trading by banks [VoxEU & IMF working paper]
Banks Will Always Suck At Trading, Badly Need A Volcker-Like Rule: Study [HuffPo / Mark Gongloff]
Banks Must Cut Deeper to Help Stock Prices, McKinsey Says [Bloomberg]
Why banks shouldn’t trade [Reuters / Felix Salmon]
2. What is the math on this? It’s like, (1) you get Libor + 10, (2) you fund at Libor, HAHAHAHAHAHA, no, I mean, fine, let’s pretend you fund at Libor, everyone else does anyway, (3) you need some capital against the loan and your capital is not cheap (at 8.5% capital ratios, 50% risk-weight, and a 10% cost of capital, that’s 42.5bps just for capital), and (4) if you’re hedging by buying 1-year CDS, which is not a great hedge on a 5-year facility, but whatever, then that eats up ~6.5bps (per Bloomberg CMA); 5-year CDS is ~32bps and so prohibitive. Perhaps there’s a way to optimize to waaaaay reduce the risk weighting there but basically you are losing money on this thing.
3. It does pay the banks several million dollars a year for borrowing insurance, though, which is still probably underpriced.
4. The study is not online; McKinsey were kind enough to send me a copy, and while they might not want me to put it up in its entirety, that excerpt gives you a good enough flavor of it that you can reconstruct the rest. Like, 52 pages of “lean solutions.”
5. I spent some time pondering this. Like, here is a thing:
That doesn’t seem like banks have systematically repriced versus corporates. But of course that comes from a world where “AA-rated banks” was synonymous with “big banks.” That world seems to be structurally gone, as ratings agencies ever so slowly catch up to reality. Related. So if it was like “once banks were AA and funded a little worse than AA corporates, and now they’re BBB and fund a lot worse than AA corporates,” fair enough.
6. D’you know, I’ve never seen that movie? Am I missing much? Is the main purpose of watching that movie in fact to be able to make Bedford Falls references when talking about A Simpler Time In Banking?
7. Reading the McKinsey recommendations to banks that they (1) reduce reliance on their balance sheets, (2) pay less, and (3) change to a less risk-oriented culture, I kind of interpreted McKinsey to be saying “hey bankers, become consultants, you’ll love it over here!”