Saturday, March 18, 2023

Silicon Valley Bank Was Uniquely Vulnerable To Intentional Asymetric Attack...,

I still want to know who started the run? Not the Peter Thiel run on March 8-9 brought on by SVB’s liquidation of long-maturity Treasuries and MBS and failed equity offering. Why was SVB already having to raise so much cash before March 8-9?

SVB’s 12/31/22 SEC Form 10K (now locked from public view) disclosed a huge run-up in time deposits during 2022. Page 81 of SVB’s 10K showed a nearly 5-to-1 imbalance in “non-U.S. time deposits” exceeding the FDIC limit, most having a maturity of 3 months or less.

Who were these “non-U.S.” depositors? Were they responsible for SVB’s sudden 2023 need for liquidity?

Why was SVB’s management team unable to understand the risk profile of this sudden influx of foreign time deposits — rather than local “parked” VC investment — and match their own investments to them (simple incompetence and the absence of a risk officer likely explains this part of the puzzle)?

Were the “non-U.S.” depositors state actors aware of SVB’s reckless and corrupt exemption from the Basel rules? 

Was the pre-March 8-9 an asymmetrical introduction of financial contagion and crisis into the U.S. banking system by outside state actors aware of the hubristic lack of regulation and oversight in the U.S. financial system?

Portfolio companies were forced to keep their deposits at SVB by their VC investors (which is… unusual) or by the terms of any SVB lending that they drew (which would be perfectly reasonable). In either case, it makes the depositor more of a victim since, once they took the VC investment of SVB loan, they didn’t have any banking risk options other than SVB. We just don’t know what percentage of depositors were compelled either way but it doesn’t matter because the more important players in the failure are the funds themselves.

This is the real scandal at SVB: how its managers’ and fund clients’ greed drove it into the ground in the last eight years of the ZIRP and pandemic boom, having spent the first thirty years building a solid business lending primarily to real businesses (albeit VC-backed). It is the shift from industrial to financial capitalism personified and on fast forward.

In 2015 50% of its loans were to portfolio companies and 33% to funds and 66% of its high quality liquid assets were available-for-sale securities, i.e. marked to market. SVB doesn’t publish a breakdown of depositors, but it would be reasonable to assume the split broadly follows the loan book, given the relationship banking approach and the fact that a the bank only takes deposits to cultivate a borrower.

By 2022, only 23% of loans were to companies and 56% were to funds. This is the killer change, if the deposit base mix followed suit. Deposits from companies are comparatively sticky, given lending relationship and other services like card merchant services etc. Deposits from funds are not sticky: the general partner is borrowing cash today (to accelerate investment pace) against an agreed schedule of future capital injections by the fund’s limited partners and, given these relationships are contractual, the only practical security for the loan is that the capital calls are paid into a nominated SVB account from which the loan is satisfied (1).

Worse, by 2022, the high quality liquid assets were only 22% available-for-sale (I.e. marked to market), down from 66%. The rest were held at book value.

So the stage is set. The overall asset base has increased by 400% but high quality liquid assets available for sale have lagged and only increased by 50%. The majority of the loan book is now lent to funds as hot money advances on capital calls and these funds (or the funds plus their puppet portfolio companies) are the likely majority of the deposit base. Cue a rumor among the herd mentality funds and that deposit base flees overnight, as the general partners pull their money and order the portfolio companies to do likewise.

If SVB had kept to its 2015 ratios, this would not have happened. The AFS losses would have forced an earlier capital increase and the deposit base would have been stickier because the portfolio companies would have been taking their banking decisions. Similarly, by chasing the funds as the source of loan growth and relying on the fund relationship to drag portfolio companies in for deposit base growth, SVB put its deposit base in the hands of a tiny coterie of people who promptly crashed their own bank….

I still think we should not approach SVB with Schadenfreude just because VC in the last decade has been unsympathetically triumphalist and Hobbesian (Uber, Wework, Palantir etc). But, having reviewed the numbers, I have revised my impression of SVB “doing God’s work” as a banker to startups. It has instead been a greedy enabler of a clique of general partner assholes.

Unfortunately, the portfolio companies have been used as a human shield by the VCs and they have not gotten what they deserved (2).

(1) The money advanced can likely be moved without penalty – if there was any requirement to hold the loan advances with SVB, this cannot be a very hard requirement because the fund is borrowing precisely to invest the money rapidly. I wonder if the funds also promised SVB that their portfolio companies would keep the investment proceeds at SVB, hence the compulsion from the funds to the portfolio companies…?

(2) revenge is a dish best served cold. The funds that borrowed their future capital to bet it all on black in the 2020-2021 peak will have torched their entire fund’s investment returns for good, given the active investment period us typical four of the ten fund years and they borrowed money from all four of those years to spray it around in two. So they will get a comeuppance – but they will probably raise a new fund because there seem to be no penalties for failure at the top in public life any more….

The idea that the deposit mix followed or somewhat followed their loan mix had not occurred to me. That generally makes sense. But recall also they bought Boston Private Bank and got with that (and presumably also solicited) wealthy individuals.