Why I Added To Silvergate After 45.33% Crash (SI)

It is a terrible feeling to be long Silvergate (NYSE:SI) and see it collapse ~49% in a day after an extended period of downward momentum, and today isn’t looking too good either.

The collapse happened because Silvergate issued a business update admitting to losing 70% of its deposits.

Everyone knew they were losing deposits, but I don’t think anyone expected them to lose this many. A downward correction seems more than appropriate.

It always feels ridiculous to defend a company that’s coming out with bad news like this and where there’s still more bad news (opinions can differ on how bad it will get).

These are times that will be interesting to reflect on three years from now. Will Silvergate be significantly higher than it is today, will it go to zero, or will it continue to languish around this level?

If I try to distance myself a little bit from the outrage and the chaos at hand and take a long-term view, my view is still that this will be significantly higher in three years. I recognize the challenges ahead and understand it may not work out if regulators uncover wrongdoing or if regulators otherwise undermine the future of the business.

One thing I believe is essential is that I understand Silvergate always encouraged customers (primarily digital asset companies) to take deposits elsewhere if they weren’t actively deploying the capital. Silvergate wasn’t paying any interest on digital asset customer deposits and therefore, the capital must be used in a way that’s more productive than the short-term yield customers could have gotten elsewhere.

Their deposit base also had a volatile history going up and down fairly wildly. They prepared for this with an excellent balance sheet loaded with liquid government-guaranteed or backed securities.

Customers like the services Silvergate provides and see value in the SEN network (this allows them to trade dollars with each other). With crypto firms blowing up left and right and all kinds of rumors being circulated about Silvergate (go on Twitter and search for $SI) there, customers may have gotten scared. Enterprise customers tend to deposit large sums, and FDIC insurance doesn’t cover that. I expected enterprise customers wouldn’t be easily scared away and trading firm deposits would be very sticky (chaos in the crypto world means great trading opportunities for arbs and market maker-type firms). I was wrong. 70% of deposits got pulled.

Normally, it is a disaster for a bank to lose deposits. Once you lose them, it is tough to get them back. You have to entice them with high yields but that sort of defeats the purpose. A bank ideally wants low-cost sticky deposits. But Silvergate’s case is fairly unique. Customers aren’t here because of the yield or because they know the brand… They are here because of the SEN network and the specialization in providing bank services to digital asset customers. This is still hard to find.

I think that if it becomes clear Silvergate is completely sound, deposits could return quickly. With this latest PR, it should be clear Silvergate is sound. The bank has cash on hand that more than covers the digital asset customers’ deposits. The latest deposit figure was slightly higher than the absolute bottom, which may (or may not) be indication deposits are starting to bounce back. I think the remaining deposits will be the most sticky. These are likely the customers that derive the most use from Silvergate’s services.

If you don’t believe in a future where crypto plays any role, this company doesn’t make sense. Ultimately, I believe digital assets will be a thing of the future, and that’s why I continue to see value here.

Yesterday’s press release contained a lot of interesting information. The company also held a lengthy call, worth listening to, where they provided a lot of additional color.

At the risk of sounding like a delusional bull, I’ll review some of the “positive” takeaways from the business update. Quotes are directly from the PR or the call transcript and my discussion follows in between:

The Silvergate Exchange Network Platform continues to operate 24/7 with average daily volume totaling $1.3B in Q4 vs. $1.2B in Q3.

The increase in volume appears small. However, we should take into account that deposits dropped by 70%. This looks like trading volume, done by the remaining customers, exploded.

This is in line with my expectation that the remaining customers derive a lot of value from the SEN-network and are more likely to be sticky customers.

Cash and cash equivalents as of Dec. 31, 2022 was ~$4.6B, in excess of deposits from digital asset customers.

It should be clear to customers that their deposits are covered, and they can pull their money anytime.

During Q4, Silvergate (SI) sold $5.2B of debt securities for cash proceeds, resulting in a loss on the sale of securities and related derivatives of $718M during Q4 2022.

It looks like this loss is primarily losses on the bond portfolio that were already present before this run on deposits. It is just that selling assets to cover redemptions forces the company to recognize these losses within its financials instead of just pointing them out in the discussion in the 10-K. It is mostly the result of the Fed raising rates and banks (including Silvergate) taking a hit on the long end of their existing bond portfolios.

Accountancy rules allow you to categorize these loans/bonds as held-to-maturity or HTM, meaning you’re assuming you’ll get paid back in full at the end of their life. If you can’t realistically apply HTM, like when 70% of your deposits get pulled, you have to mark these to market. However, the economic loss already existed in reality whether accountancy reflects it or it is only reflected in the footnotes.

Here is some color from the call:

... We did make the decision late in the fourth quarter that we no longer have the ability and the intent to hold the securities that were previously categorized as held-to-maturity. And that was fairly obvious given the fact that we were selling a significant portion of our securities portfolio. And as I mentioned in my prepared comments, we will likely sell additional securities here early in 2023 to further pay down wholesale borrowings.

And so just to make sure nobody is double-counting here, you could go back and look at the mark on the available sale portfolio at September 30, and then also look at the mark on the HTM portfolio, which was disclosed in our 10-Q. And you could assume that all of that now is — has either been realized through that $700-plus million of loss and/or is included in the remaining roughly $300 million of mark on the portfolio…

Another comment from the call I liked is the following:

At December 31, 2022, the company held $5.6 billion of total debt securities at fair value, all of which are U.S. government or agency-backed and available for sale, and which include unrealized losses of approximately $0.3 billion.

The company anticipates selling a portion of these securities in early 2023 to reduce wholesale borrowings, which will result in recognition of a fourth quarter impairment charge related to the unrealized loss on those securities expected to be sold. As we always have, we will continue to evaluate our balance sheet and liquidity management needs, which will depend on deposit flows and customer behavior.

The company is anticipating paying back wholesale deposits. To my understanding, they’ve been raising deposits (to decrease the need to sell securities) from nondigital asset customers by enticing them with high rates. Of course, this is very high-cost funding, and they would ideally get rid of it under more stable conditions. This sort of suggests they expect more stable conditions for next quarter already, AND they’re not saying they expect selling to meet digital asset redemptions.

The last comment from the call I wanted to highlight is the following(emphasis mine):

So as Alan characterized, there was a crisis of confidence and a lack of trust in the industry that happened in the fourth quarter. And so we had clients that were proprietary traders, market makers that had been doing business with each other for sometimes 6, 8 years, that just stop doing business with each other and pulled their — essentially pulled all their deposits. We had some clients that moved — these are crypto-native firms that moved almost completely into U.S. treasuries.

So that was sort of the dynamic that happened in the fourth quarter. As we were talking with our clients and asking them, “Hey, when might you shift to a risk-on position?” they really couldn’t tell us. That said, we didn’t have any clients that said that they were exiting the space altogether. Perhaps there will be some that do, but we didn’t receive that feedback. And our clients were generally supportive of Silvergate despite the fact that they pulled their deposits. And just really given the overall circumstances, decided to take that action. And — but seemed to be committed to the space and willing to come back when market conditions are right.

This client feedback, obviously provided by the company (hence with a bullish bias), is a confirmation of my hunch that the deposits are more likely to return than if a similar amount of deposits are left at a “normal” bank.

If you believe crypto is toxic or uninvestable, I can’t make any case for this bank that makes sense. I don’t love buying the bank on a book value basis. Although the deposit news was terrible, the business update gives clarity and survived a tremendous run. I saw an article in the FT today saying:

It has simply decided to retrench for the time being. Unlike others involved with cryptocurrency, it may find a means to survive financially. Its reputation though will not.

I can’t entirely agree its reputation will not survive. I would say its reputation should be enhanced, given the bank run it endured. In my opinion, a run that wasn’t entirely necessary but encouraged on social media.

In the short term, the pain will likely continue for a while. Earnings power is unlikely to “normalize” in 2023 and maybe not even in 2024. However, I do see paths where the deposits return, and the bank goes back to making $3+ per year in EPS while growing. In the shorter term, I can see paths where it gets to a pre-2020 level of $1-$1.20 per share. That is likely enough to support the share price as is, without any growth. Right now, people all talk about survival and book value (as they did after 08′ about every company anywhere).

Me too. But the crypto world moves fast, and if we’re ever getting past this deleveraging, fraud shakeout, people start thinking again about the embedded optionality. About the value of future earnings. I added shares yesterday between $12.25 and $13.90. On the Special-Situations report, I’ve included an admittedly speculative option position that could be interesting to bullish investors.