A crisis we don’t need: How not to cause a bank run

Mar 10

By Peter Kolchinsky, Tess Cameron, and Alex Martinez-Forte

Peter Kolchinsky is a founder and Managing Partner at RA Capital Management and author of The Great American Drug Deal. Tess Cameron is a Principal in Strategic Finance at RA Capital Management. Alex Martinez-Forte is an Analyst, Strategic Finance at RA Capital Management.

March 10, 2023

To our friends, peers, and fellow board members of companies with deposits at any bank,

While we thought we might be able to stem the tide, the panic around SVB’s solvency has run its course resulting in the FDIC receivership of SVB. The primary impact to companies seeking to withdraw deposits is likely to be a delay in getting their cash (FDIC has said that banking activity will resume on Monday, March 13th). We expect that anyone that didn’t have another account with a bank is opening one now. We are now working with companies on a range of solutions to the problems (e.g., making payroll) stemming from SVB’s failure.

Though the panic has taken SVB as its first victim and is, at worst, now an inconvenience for many of its clients, it needn’t get worse. Don’t participate in the contagion.

Let's acknowledge that this panic is entirely self-made. Yesterday, SVB had a strong balance sheet, one of the lowest loan-to-deposit ratios in the banking sector, and was prudently planning to bolster their balance sheet with equity after taking a loss on some of their most liquid securities. While SVB shareholders and employees will absorb losses, we expect depositors will be made whole (as with most bank failures in the past)(1). SVB’s deposits represent about 83% of their assets (similar to other venture banks), and these assets are fairly conservative, marketable, and straightforward to borrow against.

If you are wondering whether to diversify away from the bank you are now with (if other than SVB), then know that the time to do that is almost certainly not now precisely because everyone else is having the same thought. If you do, you are participating in contagion and contributing to creating a problem that doesn’t exist. Diversification is a worthy goal, but it also has to be paired with a resistance to participating in runs on banks.

Consider how things started for SVB: some companies saw news about the equity raise and started asking “will my deposits be okay?” and thinking “maybe, just to be safe, I should move my money to another bank.” When other companies heard that, they thought “well, I certainly don’t want to be the last one in line at the ATM. Maybe I should move my money, too, before it runs out.” Everyone started worrying everyone else would move their money. These are the beginnings of a classic “run on the bank,” and, indeed, our sector succumbed to mass hysteria.

At the first signs of panic, the prudent thing for everyone else to have done was quell it. Instead, we heard false rumors upon true rumors until everyone was convinced that everyone else was withdrawing their deposits and all the rumors became true.

In a few weeks, we’ll look back and see that this panic was pointless – a rumor-driven distraction from more pressing business. Our urge for everyone to keep calm is now intended to prevent further self-inflicted harm from companies panicking about whether they should turn the SVB run into a game of musical chairs across the rest of the banking industry.

No way to live

Though we are urging everyone to resist participating in a bank run, we need to acknowledge that there is an advantage to being early in a bank run if you’re smart or lucky enough to be early. If you call early, you’ll likely be able to get all your money out of that bank and won’t face risks of having your payroll delayed.

Companies that called SVB on March 9th at 9am probably got their money transferred that day to another bank. Companies that called at 4pm probably did not. It may feel like one’s fiduciary duty is to always be the 9am call to get your money out first. But if we are always living trying to get the insight that makes us the 9am call, then we will be spending our lives on the edge of our seats worried about every change in a bank’s balance sheet. It’s no way to live. We have real risks to manage inherent to innovation without creating the conditions for dealing with banking risk.

Some might think that diversification is a solution. Spread your deposits across five banks and you’re all set. But no, because the moment there is a rumor about Bank #1, you’ll panic and transfer your money to the other banks while others worry about risks to other banks. Suddenly, you’re caught up in five bank runs. No bank is immune from a bank run. The only way to avoid them is for everyone to resist the urge to panic.

Diversification is good in case a bank actually does run into problems so that a company can meet payroll, but diversification only works if panic doesn’t spread.

So rather than live your life hoping to always be in a position to lead a bank run with a 9am call to your bank and probably end up being the 4pm call, a far saner way to live is for all of us to recognize that there’s no need for anyone to be the laggard in a bank run if we just don’t start them.

A note to bankers

We also witnessed rival bankers call companies, encouraging them to switch their business away from SVB. We respect competition but urge caution: stoking a run on a competitor may spread the flames too close to home. Panics are not logical. They are not your friend. They will burn everyone alike.

A more mature message would have been: “We would love to win your business because our service is better/our rates are more competitive, but we do not want your business just because you are worried that SVB might be insolvent; it would be dishonest of me to tell you that your deposits are in any real danger. The reality is that SVB will in all likelihood honor your deposits, one way or another. So don’t just switch if it’s out of fear for your deposits.”

A banking system that resists contagion is important to all banks; if and how they coordinate conveying that message will be interesting to watch.

Evolution of this call for calm

Below is a letter that we drafted late last night (March 9th) and were preparing to send out this morning (March 10th) until we saw news that SVB halted its stock and decided to wait for more information. In the next couple of hours, it became clear that nothing was going to save SVB from whatever its fate would be, and we then amended this piece to include the broader message above urging against letting contagion take root.

The original letter below still presents a useful analysis of SVB’s financials that explains why a run on SVB was irrational. And the arguments here are still valid, not that they will change the fact that the run on SVB has already happened. It took just less than a day for rumors, many of them misinformed and just plain wrong, to turn into a stampede.

So please read this while considering how the same letter could be written about any of the top banks that one might consider as an alternative to SVB. Because no bank is safe from a bank run. If all the depositors of any of the most credible and well capitalized banks in the world (e.g., SVB) decide to panic, they will create their own crisis. Panic is pointless. Instead, each company should look ahead. Study the banks you work with. If you see a problem, diversify or switch. But don’t do it just because you see others doing it; at that point the panic is the problem, not the balance sheet. The remedy is not to switch but to stay calm and urge others to stay calm. We need not act like a frightened herd.

* * * * *

Original letter written late on March 9th:

A crisis we don’t need: Much ado about SVB Solvency

To our friends, peers, and fellow board members of companies with deposits at SVB,

Management teams and boards across the biotech and tech sectors are deliberating whether to join in the wave of companies that seem to want to pull their money out of SVB. Most don’t even know what’s going on with SVB or what to be scared of exactly but that’s the nature of a “run on the bank”… people just run because they don’t want to find out what’s chasing them.

The trouble is that the run is often the problem… It's like a herd on a plateau that hears a pebble fall, thinks the plateau is disintegrating, and runs off one edge into the abyss.

It would help to figure out if the plateau is really disintegrating.

What people want to know is “if I don’t pull my money now, will SVB be able to give it to me tomorrow or the next day or next week or next month when I need to make payroll?” To answer that, it helps to understand just how much cash SVB has and can access.

Here’s how we break it down and why we see no cause for panic.

How did we get to this moment?

SVB provides commercial banking services to tech and healthcare companies. Like every bank, they make money primarily by charging interest on loans for more money than they pay deposits.

Amidst the downturn, companies are raising less and spending less. But they are spending more money than they are raising. This means that the balance in their bank accounts is going down. Banks need a deposit base to lend against (on a bank balance sheet, deposits are liabilities and the loans are assets). In 4Q22, SVB thought they were through the worst of the deposit outflows and told investors they expected net outflows of deposits to slow down.

But the downturn continued into 1Q23, which meant that SVB decided to restructure a portion of their investments to be more liquid (in the event they need to use them to cover deposits) and earn higher interest. This sale of almost all of their available for-sale securities triggered a loss (since treasuries have lost value as interest rates rise), and on March 8th SVB announced plans to further bolster their balance sheet with a $2.25B raise of equity and debt. Some would consider this prudent.

However, on the morning of March 9, some companies saw this and started asking “will my deposits be okay?” and “maybe, just to be safe, I should move my money to another bank.” When other companies heard that, they thought “well, I certainly don’t want to be the last one in line at the ATM. Maybe I should move my money, too, before it runs out.” Everyone started worrying everyone else would move their money. These are the beginnings of a classic “run on the bank” but that panic can be quelled if everyone can be dissuaded from succumbing to mass hysteria, which is the purpose of this letter.

Cash is there… Not all at once, but who even needs it all at once?

If SVB were Lehman Bros and held nonsense assets, that would be a problem. But SVB is a comparatively conservative institution.

SVB’s loan-to-deposit ratio, a key measure for a bank’s leverage, stands at about ~43%, much lower than the >70% typical of large banks. This means that SVB is well positioned to cover loan losses in the event of increased defaults. Their deposits stand at ~80% of total assets(2), meaning if everyone wanted their money back tomorrow, they would get it as fast as SVB were able to liquidate or borrow against ~80% of their assets. This is in-line with other venture bank peers (PacWest, Comerica, First Republic).

So a run on SVB where everyone wanted their money immediately would result in clients getting their money even if not all of them get all of their money immediately. And if that left clients disappointed, it would only be because the idea of everyone getting their money back immediately was never part of the tacit “contract” that a bank has with its clients. It’s a totally unrealistic expectation.

That’s because a bank is a collective enterprise, akin to a community. A community can allow anyone to take a few weeks off work, but not everyone at once. If everyone tries to relax at once, no one can since the water stops running and electricity stops flowing and the garbage piles up.

So everyone can’t get their money out of a bank at the same time and it’s foolish to try. For that, you would want a vault, not a bank. Some people mistake a bank for a vault. It’s not. A vault would not pay any interest on deposits. It also wouldn’t offer loans. Banks are a source of both liquidity and return precisely because they diversify everyone’s cash and liquidity needs. Clients tacitly agree to go about their business as usual and trust that the money will be there when they need it.

If every client keeps about a year’s worth of burn with SVB, then providing everyone with access to their cash is straightforward. If everyone invents a reason to panic and runs to withdraw their money, they stampede into a crisis of their own making. Even then, the crisis will merely be delays in accessing all of one’s cash.

All the money will ultimately be there. Just don’t ask for more than you need and enough will be there whenever you need it. That the promise of a well-run bank and SVB has all the hallmarks of a well-run bank.

The hallmarks and metrics of a well-run bank

The fact that they decided to raise money despite how dilutive that would be is an act of prudence on their part. Their equity holders will pay the price for the investment mistakes they made, but that need not be of any consequence to the hundreds of tech, biotech, and other growth companies holding deposits at SVB. The fact that equity investors are willing to buy into SVB is precisely what should reassure clients that someone is willing to take way more risk on SVB than depositors do, and therefore there is no need for anyone to run on the bank.

The rational thing for everyone to do is stop being afraid, at least for a day, and consider the numbers below, which are primarily from SVB’s strategic update on March 8, 2023.

SVB has $165B in deposits and has access to ~$180B from selling or borrowing against its assets (e.g. treasuries, short term loans, long term loans such as the venture debt SVB is known for issuing). We estimate that they would be able to at least quickly access ~$95B in liquidity quickly, of which:

$50B from borrowing against its held-to-maturity (HTM) securities(3) (i.e., its portfolio of treasuries and other debt products, that are government backed), of which they have $95B so we’re being conservative by assuming that they would start by borrowing against only 50% of the value of those securities.
$25B from liquidating of sellable securities (securitizations, CD’s, bonds, debt, etc.).
$15B from either using or borrowing against their own cash.
$2.25B from financing activities (i.e. the funds they announced they are raising now which sparked this whole tempest in a bottle).
$1.5-2.5B from held-to-maturity cash flows per quarter.

This provides a cushion of roughly $95B in the short term, equal to 58% of their deposits. If they had to, we estimate SVB management would be able to access $75B more cash to cover the remaining deposits (now we are at $170B, vs. $165B in deposits). These actions might incur heavier losses and have a more significant impact on Net Interest Income than the actions above, so we expect management would only pursue these if they had to:

~$60B from shifting off balance sheet assets to their balance sheet (i.e., their sweep, which are portions of customer deposits held at other banks).(4)
$15B from borrowing against their remaining held-to-maturity securities - this is likely the last leg of their $95B HTM securities that they would be able to withdraw and they would only do it as a last resort effort because it would fully max out their borrow capacity.

Finally, we estimate that SVB could generate a further ~$30B to fund deposits by letting some of their existing loans expire without rolling them over (which would bring us up to $200B, vs. $165B in deposits). Out of their total outstanding loans of $74B, the majority of them (56%, ~$40B) are to PE/VC funds as line of credits for capital commitments.(5) These loans are backed by LP cash and in addition to having very low credit risk, which are often <6 months in duration. We estimate that SVB will receive ~$30B(6) within three months from maturities across its loan portfolio based on their FFIEC 041 filing.

None of the above actions are good for SVB equity holders, since shareholders want to see assets generating net interest income for them, not funding withdrawals. So, while we may yet see more pressure on the stock as SVB equity investors calibrate on the impact of withdrawals on the bank’s earning potential, depositors should take comfort in SVB’s strong balance sheet.

We might wake up one day soon to hear that everyone’s collective panic has motivated the Fed to step in, or resulted in a low enough stock price for an acquirer to swoop in [post-script: or FDIC takes it over]. These possibilities provide the potential for support beyond just SVB’s strong balance sheet and is yet another reason depositors can keep calm.

The ecosystem perspective and a movie recommendation

We - healthcare, tech, and growth investors - are all better off with SVB in a strong position to compete for our companies’ banking business, as they have been for the past 40 years. SVB is a part of our ecosystem and has run itself prudently. If every company and board is only looking out for themselves, they may conclude that pulling money from SVB is the right way to mitigate risk. But if every company also considers their place in the ecosystem they should conclude just the opposite. Let’s not cause a self-induced panic that will only harm our industry.

If our call for biotech leaders to recognize that we are all part of a single ecosystem and have a duty to preserve it sounds familiar, that’s because we have made this call before, such as drug pricing policy. Some CEOs and board members think that “community stuff” is for others to burden themselves with… that they should just look out for their companies. But myopic self-interest is unbecoming of an executive or board member.

When we fail to see how much of our success depends on behaving as members of a community, we all lose. Leadership means seeing the bigger picture and finding a way to collaborate when that results in a better outcome for everyone.

In this case, collaborating means simple everyone taking the day off from freaking out and maybe watching “A Beautiful Mind” about John Nash, a game theorist for whom the concept of a Nash Equilibrium is named; this refers to the idea that sometimes people looking out just for themselves can’t help but to undermine themselves and end up with a worse outcome than if they had just communicated and collaborated.

A run on a bank with SVB’s financials is a truly dumb Nash equilibrium. If enough companies just paused, the crisis would vanish. It’s not even essential that all companies pause. Some can go ahead and continue panicking… if it’s a small fraction of SVB’s book of business, they will get their money, incur switching costs, and everyone else won’t feel a thing.

SVB is currently in a quiet period as they work out a solution. This means that communication from the company may be limited over the coming days. Let’s not fill the void with rumor mongering that only leaves us all worse off.

Sincerely,

Peter Kolchinsky
RA Capital Management, LP(7)

Tess Cameron
RA Capital Management, LP

Alex Martinez-Forte
RA Capital Management, LP

Co-signers below reviewed this letter and agree with its call for everyone to give serious consideration to just hitting pause on the panic, staying cool, and anchor to the robustness of the balance sheets of the banks with which they do business.

Please fill out the form below the list to co-sign this letter.

Below is the full list of those who have co-signed

(1) The largest comparable to SVB failures was Washington Mutual in 2008 with $188B of deposits; depositors were made whole through an FDIC-facilitated acquisition by JPMorgan Chase

(2) $165B of deposits, a decline of $9B vs. 4Q22, vs. total assets of $203B, assuming a similar decline of $9B vs. 4Q22 assets of $212B

(3) Some on Twitter have expressed concerns that SVB may be forced to sell these HTM securities and incur a loss, which would make it harder for SVB to fund withdrawals. But SVB would not do this; they would simply borrow against these securities and still hold them to maturity so that they could realize the full par value at maturity.

(4) Moving the cash held from other banks into its own deposit account would force SVB to have to pay a yield to account holders, maybe ~4.5%. So while it’s a source of cash, it would be relatively expensive and impact their Net Interest Income.  We haircut SVB’s $73B estimate to $60B assuming that a portion of their sweep deposits have been withdrawn.

(5) SVB’s loan portfolio is rather unique because 56% of it (~$40B) is in short-dated loans to Private Equity and Venture Capital firms (PE/VC) that are often <6 months in duration. What this means is that a PE fund might borrow $100M from SVB on short notice and then take its time calling that capital from its LPs. It’s extremely low risk lending but it smooths out a fund’s access to cash so that it can make many investments each quarter but call capital from its LPs in lump sums each quarter to cover multiple investments. If funds had to call capital from their LPs for each investment, LPs would burn out from the constant hassle of wiring money. If it had to, SVB could borrow against those loans to repay clients seeking to withdraw their deposits. It could also just stop issuing new loans to preserve cash for depositors withdrawing their money, but that would impact their reputation as a lender and cut into their future profits. Companies might think “who cares? I just want my money!”, but that’s myopic. Companies should want SVB to stay in business and keep generating a return from loans to VC and PE forms because it’s this kind of “other” profit stream that allows SVB to offer competitive loans to biotech companies and interest on deposits.

(6) We estimated that SVB could receive ~$30B from the maturity of loans that it previously originated. Our estimate is based on the reported $43B in loans maturing within 1 year from their 2022 FFIEC 041 filing. They also stated that 80% of loans that mature and/or reprice within 3 months are $59B in the same report. We assumed that about ~70% ($30B) of the reported $43B would mature within 3 months to be conservative.

(7) RA Capital Management, LP does not have an account with SVB though many of its portfolio companies do.

Please click here for important RA Capital disclosures.