I like reading banks' research reports on other banks these days because they give off a certain the-call-is-coming-from-inside-the-house vibe; you imagine the analyst running the numbers, looking them over, and saying "my God, this can't be right, can it? This seems to say ... I'm fired?" JPMorgan's analysts maybe suffer from this less than most but it still imparts a certain tension to the marvelous, strange, 100-page research note out of J.P. Morgan Cazenove today about global investment banks.* There are two big important points** which are:
(1) European banks are pretty pretty aggressive with how they risk-weight their risk-weighted assets, especially compared to US banks. Basel's Standards Implementation Group is moving in the direction of requiring convergence on RWA measurement, and JPM thinks that that will lead to the European banks having to revise their RWA measurements - meaning that those banks' capital positions will look much worse than they do now and they will need to shed RWAs and/or raise capital.
(2) You can quantify the return-on-equity effects of new banking regulation - including Basel RWA convergence, but also things like derivatives clearing, the Volcker Rule, etc. - on the big global banks, and those effects are bad. Bad for shareholders, anyway: per JPMorgan, global-bank average ROE would be 16% in 2013 but for those regulations, while after giving effect to them it will be just 6.3%.
But I presume that like any good utility maximizer you care only about your comp, so the important takesaways are (1) 6.3% is not good enough and (2) it will be remediated out of your pocket. Which leads JPMorgan into the truly chilling:
Per JPMorgan, regulations that are being phased in will cause 28,241 additional layoffs just among the eight big banks named here - ignoring in particular more banking-focused US names like BAC, C and of course JPM itself - and it will cause a $70k average drop in comp for those who are left, all just to get back to a plausible run-rate of 13% (for "Tier I" DB/GS/Barclays) or 10% (for "Tier II" others) ROEs. So vote Romney, JPMorgan almost says.
I don't really know what to make of this. I guess you have to read it together with this:
As record-low interest rates limit returns on assets, banks have become more focused than ever on funding costs. That’s led them to lean more on the cheapest form of borrowing available: federally insured deposits.
Bank of America pays about $500 million a quarter in interest for its $1 trillion of deposits compared with about $2.5 billion for $300 billion of long-term debt, CEO Brian T. Moynihan, 52, said on a July 18 investor call. ...
Goldman Sachs has more than doubled its deposit base to $57 billion since 2008 and wants to raise more because the cost of three-year deposits is about 200 basis points, or 2 percentage points, fewer than issuing three-year debt in the bond market, Treasurer Elizabeth Beshel Robinson said July 24.
So I guess the tradeoff for all the regulations imposed on banks for being systemically important is the cheaper funding that they get by being systemically important. Some of this is meh; let's say GS saves about 2% of 50% of $57bn from living in a post-2008, deposit-funded world; that looks like $570mm to me, which if it goes 25.1% to comp saves ... hmm, about 10% of those 4,400 lost jobs (at $314k a year). I guess BofA does better? And MS worse? But really you have to assume some sort of systemic risk benefit across the board, as your wholesale funding is also cheaper:
Jefferies’s bond yields show that it costs the firm about 5.81 percent to borrow until 2021. That’s more than the 4.38 percent yield on 2022 debt issued by Goldman Sachs. While Bank of America has the same Baa2 credit rating from Moody’s Investors Service as Jefferies -- two levels below Goldman Sachs’s A3 rating -- its 2022 bonds yield 3.99 percent.
I rough out that if all of GS's $948bn balance sheet is about 58bps cheaper than it would otherwise be, that saves all those lost $314k/year jobs. Or, put another way, if GS got out of that regulatory regime at the cost of more expensive funding, it'd probably be losing those jobs anyway.***
Anyway! Are these numbers of the "banking will be much smaller, less profitable, less risky and more utility-like" school, or are they not? Sort of in the middle, no? The days of $600K+ average comp and 20% ROEs may be long past, but JPMorgan's estimates are compromisey: equity investors will still get double-digit returns, while average comp will go back to the levels of a decade ago,**** though it will be paid to a continually dwindling pool of bankers. So there will be pain, but not enough, perhaps, to change the overall character of banking. As JPMorgan sort of acknowledges: their estimates of the costs are "ex regulatory arbitrage." I have high hopes for that.
* That I am not putting here in its entirety out of vague respect for the proprietariness of bank research. One day I will run an experiment on this. For now here is JPM's math on the ROE effects; the RWA math is less susceptible to excerpting and more (?) hocus-pocus anyway:
** And one wee little one which is downgrading GS to underweight on valuation. Note that JPM Cazenove upgraded GS to overweight on July 6, 2010 and, y'know, yay. DISCLOSURE: LONG GS UNTIL I SLANDER THEM SUFFICIENTLY TO LOSE MY RSUS.
*** I mean, this is pretty lazy. Some of the regulatory items JPMorgan names are specific to being a global diversified bank, or even a deposit-taking one (Volcker), but some aren't and apply equally to Jefferies - things like derivatives clearing, etc..
**** Actually lower than that; did you know GS's average comp was $340K-ish in 2002? 2002!