Regulatory implications from the demise of Silvergate
by Ryan Shea
- Cryptocurrencies have been impacted by macro headwinds and continued fallout from last year's cryptowinter as Silvergate – a crypto-friendly tradfi bank – just announced it would “voluntarily liquidate” its assets and wind down operations.
- The troubles that led to Silvergate’s demise, although triggered by the failure of FTX last November, were not crypto native. It was an age-old tradfi banking problem: borrowing short - lending long.
- Nevertheless, it will be viewed as validating US regulators' concerns about crypto firms and the potential negative spillovers into tradfi.
- SPDIs are one viable solution. The State of Wyoming has chartered four such entities, one of which - Kraken Bank - is set to launch “very soon”.
- However, at the Federal level SPDIs are not being welcomed with open arms as Custodia – one of the four SPDIs – found out recently when the Fed denied their application to become a member of the Federal Reserve System.
- Without access to US tradfi banking services, crypto firms will simply be forced off-shore or to use alternate methods to conduct fiat on-off ramping. It will also fail to deliver what US regulators are expecting: insulating tradfi from crypto.
While still up handsomely year-to-date (30%), since late February cryptocurrency prices have struggled with the market leaders Bitcoin and Ethereum giving back roughly a third of their gains. In part, this weakness has been driven by ongoing macro headwinds (see below). However, the crypto industry has also been impacted by continued fallout from last year’s cryptowinter.
Late last month Silvergate, a regulated US financial institution that provides banking services to crypto businesses, revealed it was delaying its annual 10-K filing as it required additional time to evaluate its financial reports and to allow independent auditors to complete certain audit procedures. It also warned that it may end up less than well-capitalized meaning its ability to continue as a going concern may be affected. A few days later, and Silvergate just announced that it would “voluntarily liquidate” it’s assets and wind down operations. A salutary reminder that Ernest Hemingway had a point when he stated:
“How did you go bankrupt?" Two ways. Gradually, then suddenly.”
- Ernest Hemingway
A Perfect Storm
FTX’s failure was a significant catalyst to Silvergate’s fall from grace. The bankruptcy of what was the fourth largest crypto exchange globally triggered a surge in risk aversion and crypto investors scrambled to offload as much risk as possible. According to Silvergate’s Q4 earnings report, during this disorderly deleveraging Silvergate experienced a $5bn drop in its digital customer deposits to $7.3bn - a 40% drop in deposits in the space of a couple of months is a pretty nasty shock for any banking institution whether crypto-friendly or not.
Initially, Silvergate attempted to satisfy its clients demands for cash by borrowing from the San Francisco Federal Home Loan Bank (FHLB-SF). This may seem to be an odd funding source for a crypto-focused bank and it is. Most likely, the relationship between the two entities dates back to the days when Silvergate was a more vanilla bank engaged in commercial and residential real estate lending. For some still unknown reason the FHLB-SF recalled these loans leaving Silvergate with the option of either borrowing from the Fed, via the discount window, or selling off assets held on its balance sheet. Again, for some unknown reason, it decided on the latter.
Ostensibly, this should not have been problematic. Silvergate was, by regular metrics, a sound bank. At the end of September last year, before the collapse of FTX, it had a tier 1 capital ratio of 10.71% - more than double the threshold regulators deem a bank as being “well-capitalized”. Part of the reason why its tier 1 ratio was so high was because the assets held on its balance sheet were high quality, largely comprised of deposits at other banks and US government and agency bonds which get a zero risk weighting.
This risk rating reflects the view of the regulators that these are the safest financial assets. Safe, however, does not mean you can’t lose money on them. As most people understand the price of bonds (debt) is inversely related to the level of interest rates: higher interest rates mean lower bond prices and vice versa. With the Fed aggressively tightening monetary policy last year to bring down inflation, the price of US government bonds cratered. Silvergate’s need to raise cash to fund customer withdrawals meant it was a forced seller in a bear market, a situation that is rather toxic to one’s financial health as they found out when the $5.2bn debt securities they sold in Q4 resulted in losses totalling $750mn. Consequently, the bank’s tier 1 leverage ratio fell to a level barely above the 5% “well capitalized” threshold.
The March 1 news that Silvergate may no longer be a viable going concern clearly spooked the market as evidenced by the 50% slump in its stock price on the day of the statement. It also prompted many leading crypto entities – Coinbase, Circle, Paxos and Gemini – to publicly state they were no longer doing business with Silvergate. Add a more hawkish Fed combined with the increased scrutiny towards the crypto sector by US regulators into the mix and it is clear Silvergate was in the midst of a perfect storm. Little wonder, it has decided to shut up shop.
As I wrote in my previous research note, in January the Fed, FDIC and OCC issued a joint statement warning that “funding from crypto-asset-related entities may pose heightened liquidity risks to banking organizations due to the unpredictability of the scale and timing of deposit inflows and outflows ”. Silvergate’s subsequent failure is a clear demonstration of these risks materializing. Indeed, it may have been what prompted the US authorities to release the statement in the first place.
Clearly, this will have nowhere near the impact on the tradfi sector as the failure of Lehman Brothers did 15 years earlier because Silvergate is not a systemically important financial institution. Nevertheless, it is being viewed as validation of the concerns of regulators in the US, and elsewhere, of the potential for negative spillovers from crypto to fiat – see image below.
Certainly, it will not discourage what Nic Carter has dubbed Operation Choke Point 2.0: a coordinated campaign to deny crypto firms access to tradfi banking services. If anything, it will strengthen their resolve.
This is unfortunate because the troubles that led to Silvergate’s demise were not crypto native. Yes they were triggered by the failure of a crypto firm (FTX) but the problem was an age-old tradfi one of liquidity or duration mismatch in the balance sheet: borrowing short - lending long. At most one could, and probably should, criticize Silvergate’s decision to have such low cash or cash equivalents on their books given its customer base was heavily concentrated in a single asset class and whose deposits and withdrawals were always, therefore, likely to be strongly positively correlated.
Full Reserve Banking
One way to prevent a repeat of the Silvergate experience is for banks operating within the crypto space to be full reserve banks which do not lend out any customer demand deposits. Because it means funds are always available to meet customers withdrawals a bank run becomes impossible.
In 2019 the State of Wyoming introduced legislation – clearly aimed at the crypto sector - authorizing the charter of so-called SPDIs (Special Purpose Depository Institutions) to receive customer deposits, provide custody and asset management services, but which are...
“… generally prohibited from making loans with customer deposits of fiat currency and they must at all times maintain unencumbered level 1 high-quality liquid assets valued at 100% or more of their depository liabilities.”
So far the Wyoming Banking Board has approved four SPDI charters, the first of which was Kraken Bank, an affiliate of Kraken the crypto exchange. The company’s chief Legal Officer, Marco Santori – no doubt with an eye on recent developments – last week confirmed that Kraken Bank is set to launch “very soon”.
Given their structure, SPDIs should be a much better banking fit for crypto than fractional reserve banks given the seemingly high correlation between customer withdrawals and deposits. It would make the space safer for crypto users because they could be assured that the necessary fiat funds are available on-demand if they decide to off-ramp en masse. Furthermore, because crypto-focused SPDIs are part of the tradfi banking system, regulators would have full transparency on their balance sheets and be better positioned to monitor risks.
Despite these apparent benefits, at the Federal level at least, it is fair to say SPDIs are not being welcomed with open arms. On January 27, Custodia Bank – one of the four SPDIs chartered by Wyoming - finally (two years after submitting its application) learned that the Fed had denied its application to become a member of the Federal Reserve System, something that would have given access to Fedwire - the real-time settlement system used by US tradfi banks. The reasons given for its decision were:
“The firm's novel business model and proposed focus on crypto-assets presented significant safety and soundness risks.
The Board also found that Custodia's risk management framework was insufficient to address concerns regarding the heightened risks associated with its proposed crypto activities, including its ability to mitigate money laundering and terrorism financing risks.”
In response to the news, Custodia announced it would continue to pursue its lawsuit against the Fed over its application. Given other tradfi companies, such as BNY Mellon, have been given authority to provide custody services for crypto, it will be interesting to see how Custodia gets on, especially as it overcame a significant early hurdle when the court denied the defendants’ (the Federal Reserve Board of Governors and the Federal Reserve Bank of Kansas City) motion to dismiss on four of Custodia’s key claims.
Unless, by their actions, US regulators are able to pressure/persuade/convince [delete as you feel appropriate] the general public – or at least a significant portion of it – to finally give up on crypto, which is not going to happen for the reasons I outlined in the previous research note, it is hard to see much merit in the approach being taken by US regulators. Without access to US tradfi banking services, crypto firms will either be forced off-shore or to use alternate methods to conduct fiat on-off ramping.
Think about that for a second.
By seeking to limit the damage to tradfi from potential contagion from crypto US regulators not only will be giving up any potential competitive edge arising from the applications of this technology but they will also lose visibility and hence the ability to monitor what crypto firms are doing and how much of a risk they potentially poise to the US financial sector. One may be tempted to think that by cutting off crypto’s direct access to the US fiat banking system the risks would evaporate but that would be wrong. Many other nation-states are more than happy to step in and support crypto businesses and unless the US imposes draconian banking/capital controls on them crypto users will be able to find a way to on and off ramp.
One rather obvious workaround would be for crypto firms unable to conduct their business in fiat to increase their reliance on, and use of, stablecoins, such as Tether – the market leader – or USDC. This is viable as long as stablecoin issuers are able to find a fiat ramp, something that seems very do-able.
How can I say this?
Quite simply, because of the existence of money laundering in the fiat world. Governments around the world are supposed to be tackling this problem, but the flow of “dirty money” continues– see image for a stylized example - and in size. According to the UN, the estimated amount of money laundered globally in a single year is between 2-5% of global GDP or $800bn - $2tr.
If all countries around the globe adopted the same rigorous standards in implementing anti-money-laundering regulations then it would not be such a significant problem. However, as per the chart below - courtesy of the Basel Institute of Governance - this is not the case. By their metrics, much of the global south – representing a sizeable chunk of global GDP - have AML/CFT (anti-money laundering / Combating Financing of Terrorism) systems that are below average in terms of effectiveness. The global fiat money system has more ways to launder dirty money than a sieve has holes.
Moreover, while the estimated amount of money laundering is large – it exceeds the total market cap of cryptocurrencies for example – it pales into insignificance compared with the flows in the international monetary system. To provide some context, according to the latest BIS triennial FX survey, global trading in OTC FX markets reached $7.5 trillion per day in April 2022, of that 80% has a USD component – see chart. This equates to an annual turnover of $1,950 trillion (that’s a lot of zeros). Money laundering constitutes 0.1% of annual FX turnover. Those tasked with trying to identify this dirty money are looking for the veritable needle in a haystack.
Global FX Market Turnover By Currency
If governments are unable to stop dirty money entering the fiat money system, do you think they will be able to stop someone, or some entity such as a stablecoin issuer, from on/off-ramping crypto to fiat using a protocol or entity which is not within their jurisdiction or national borders?
No, me neither.
Until next time.
Ryan Shea, crypto economist
 See: https://ir.silvergate.com/news/news-details/2023/Silvergate-Capital-Corporation-Announces-Intent-to-Wind-Down-Operations-and-Voluntarily-Liquidate-Silvergate-Bank/default.aspx
 See: https://s23.q4cdn.com/615058218/files/doc_financials/2022/q4/Ex.-99.1-SI-4Q22-Earnings-Release-FINAL_v2.pdf
 Perhaps they pulled the loans because Home Loan Banks are not really supposed to be in the business of supporting crypto-friendly banks. The following article contains other well-founded speculations as to the reason why they may have pulled Silvergate’s loan – see: https://www.coppolacomment.com/2023/03/lessons-from-disaster-engulfing.html?m=1
 There is a presumed stigma from borrowing from the discount window because it could be seen as a sign of weakness - https://www.federalreserve.gov/econres/notes/feds-notes/stigma-and-the-discount-window-20171219.html
 See: https://www.stlouisfed.org/bsr/risk_weights_on_balance_sheet_assets_ffiec_041051/story_content/external_files/Risk_Weights_On_Balance_.pdf
 Silvergate expects to sell an additional $1.7bn of debt securities in Q1 and in anticipation of this recorded a $134mn impairment charge.
 See: https://www.sec.gov/Archives/edgar/data/1312109/000110465923027353/tm238251d1_nt10k.htm
 See: https://www.theblock.co/post/216548/silvergates-network-unravels-leaving-crypto-industry-in-search-of-new-partners
 See: https://www.reuters.com/markets/rates-bonds/hawkish-powell-puts-50-bp-fed-rate-hikes-back-table-2023-03-07/
 See: https://www.federalreserve.gov/newsevents/pressreleases/files/bcreg20230223a1.pdf
 See: https://www.piratewires.com/p/crypto-choke-point
 See: https://wyomingbankingdivision.wyo.gov/banks-and-trust-companies/special-purpose-depository-institutions
 The three other SPDIs are Custodia (formerly Avanti Bank), Commercium Financial and Wyoming Deposit and Transfer.
 It was initially slated to launch in 2021 - see: https://www.theblock.co/post/217177/kraken-is-on-track-to-launch-bank-very-soon-despite-regulatory-weird-place
 See: https://www.federalreserve.gov/newsevents/pressreleases/orders20230127a.htm
 I will revisit this topic shortly.
 See: https://decrypt.co/122525/custodia-lawsuit-fed-everything-need-know
 See: https://www.dwt.com/blogs/financial-services-law-advisor/2022/11/custodia-bank-digital-assets-court-wyoming
 See: https://index.baselgovernance.org/api/uploads/221004_Basel_AML_Index_2022_72cc668efb.pdf
 NB: If stablecoins do become a workaround, it does not require all transactions conducted in them to be on or off-ramped. All it requires is that there is the ability to off-ramp in order to ensure their parity valuation with the US dollar is maintained. That is to say, it would not require much crypto-to-fiat transaction volumes, making it hard to detect.