If you read a lot of media coverage of Goldman Sachs earnings you get the sense that the most important number the firm reports is average compensation per employee, which this year was a nice oh-so-close-to-round $399,506. I CONCUR, of course.1 Also of interest is the comp ratio, which was only 39% this year, as less of the spoils of Goldman's labors go to the people in the building doing the labors, and more go to the people providing the capital. Progress!
The analysts on the earnings call were not all that focused on comp, which I attribute to jealousy, but there were some exceptions. Like JPMorgan's Kian Abouhossein, who pressed the Viniar/Schwartz CFO tag-team about expenses and headcount in Investing & Lending, playing an enjoyable guessing game with the twin CFOs about staffing levels in Investing & Lending:2
I mean, there are only few hundred -- I assume there are only a few hundred people running in this division. I can't believe there's thousands of -- I would be even surprised if it's 1,000 people. So I'm just wondering why you're having $2 billion to $3 billion of expenses. Is it interest expenses or is it something else? I just don't understand why there's such a big expense level.
Because the few hundred people are paid really well? Other? Dunno. You can guess why Schwiniar might have stalled here (and on a later question about I&L Basel III RWAs); the Investing & Lending business model has gotten some negative attention recently. The problem is basically that it does things like investing and lending, which almost violate the Volcker Rule, or would if it existed, which it doesn't, yet.
Here is the FT's Tracy Alloway on Goldman's earnings:
Revenues in the period jumped 53 per cent to $9.24bn, led by a 63 per cent increase in investment banking and a 109 per cent increase in the bank’s market-making activities. In principal transactions, where Goldman makes its own long-term investments, revenues rose about 135 per cent to $1.96bn.
While the so-called Volcker rule prohibits banks from making short-term trades for their own accounts, it still allows them to make longer-term investments such as buying property or stakes in various companies. Goldman reported a $334m gain on the value of its investment in ICBC in the fourth quarter, for instance.
But of course the market-making transactions are principal transactions too; by my count a little over half of Goldman's revenues are from using its own balance sheet:3
It's just that those revenues now sort of live in two buckets:
- "Market making," which involves relatively short-term use of Goldman's balance sheet with plausible reference to client demand, and
- "Other principal transactions," which basically involves >60-day use of the balance sheet.
Over the last few years ... meh, if you squint you might see market making down and "other principal transactions" sort of stable-but-fluctuating, I dunno:
One thing we talked about when we last talked about Investing & Lending is that, from a Volcker Rule perspective, I&L is sort of a nicer business than market making. I mean sure I&L is actually prop trading, while market making actually isn't, but "short-term trading with your own balance sheet to make markets for customers" is fairly complicated to distinguish from "short-term trading with your balance sheet to take a prop position." Whereas "long-term trading with your own balance sheet to take a prop position" is easily distinguished from "short-term, um, that"; you just hold your positions for longer than 60 days and you're fine.
Is the Volcker Rule driving Goldman into longer-term prop trading and away from market making? No, probably not, mostly I'm just being annoying, though you never know since the rule isn't final yet. But I was reminded of the possibility by the JPMorgan Whale Report this morning. Here's what the report has to say about why JPMorgan's Chief Investment Office - which, like Goldman's I&L division, invests the firm's own money on a medium-to-long-term basis - could get away with murder for so long:
CIO is not a client-facing business and does not involve the host of regulatory, risk and other limits applicable to dealings between the lines of business and their clients, which require more attention from various control functions, including compliance, audit, legal and finance. There was no meaningful effort to ensure that, notwithstanding this fact, CIO was subject to appropriately rigorous risk and other limits and was updating those limits on a regular basis.
I mean, gosh, I hope Goldman's investing business has auditors. And, it being Goldman, I'm sure they're subject to rigorous risk and other limits. Still: doesn't that list of characteristics of non-client, longer-term, Volcker-permitted principal-investing businesses sound nicer than the strict control and regulatory second-guessing of a market-making business? For the people in those businesses, I mean. For the people providing the capital it may go the other way. But why would anyone care about them?
1. Nah. Among other problems is the one of false precision; Goldman reports four digits of comp ($12.94bn) but that somehow becomes six digits of average comp, at $399,506. I would spin a conspiracy of Goldman adjusting its comp so as to pay a $400k average while getting the press to report a 3 handle, but: no. Nobody cares about average comp.
2. Though he later loses points for saying "you've done a great job on reducing staffing this year." Insensitive!
3. So btw it seems sort of unfair that 3/5 of Goldman's revenues go to the shareholders who provide that balance sheet? Or is that just my bias talking?