Bank Run? A quick thread about Silicon Valley Bank (SIVB), and how this could change the entire macro picture in the near term 🧵

First, banking 101: Banks make money by taking in deposits from folks like you and me, and lending it on to others at a slightly higher rate. It's not a risk free business since on the asset side, the loan can default, and on the liability side, the deposits can go away.

So in regular times, they make the net interest margin (NIM) between borrowing and lending, as well as fees from all sorts of other services.

When banks get into trouble, classically, it's been because the loans "soured". i.e. The loans on their books are unlikely to get their money back. This has to be marked down, and since the depositors are not on the hook for the losses, it's a hit to the bank's capital.

Here's an illustrated version of how an actual loan works in terms of bank balance sheet: https://www.pennyworks.com/article/how-banks-work-and-why-they-dont

But in essence, in the simple illustration below, if the $5 of loan goes belly up, the loss is taken by the equity holders, and the bank is left with just cash to pay back depositors

Now in today's environment, banks are required to hold a lot more equity and tighten lending. Because if this, banks have had a hard time finding suitable loans to invest in. So instead of lending, they resort to buying securities such as government bonds and other assets.

This pool of assets comprises the bank's portfolio, and is split into 2 parts. The Available For Sale (AFS), vs the Held to Maturity (HTM).

It relates to accounting differences, but suffice to say banks are allowed to trade the AFS ones, and generally not the HTM.

However, in this current environment, the uniqueness is the speed with which the Fed has hiked. Many banks own huge AFS portfolios, and if they have a maturity of 3-5 years, a 4% increase in rate implies a 12-20% marked to market loss on their portfolios.

Now most of these investments are in high quality assets, so the likelihood of actual losses is low. However, bond prices have to adjust to market rates, and when rates go up a lot, prices have to come down. It eventually "pulls to par" but only at maturity.

This is essentially interest rate risk. This leads us to Silicon Valley Bank. According to the WSJ, they had an AFS portfolio that was worth ~$30B on the books, but with a $2.5B marked to market loss. They decided to sell all of it, take the hit, and raise $2.25B of equity

There's an even bigger HTM portfolio, that's worth $90B+. With an even larger $15B unrealized loss. But the way that book works is that as long as you don't sell it, you'll get your money back, it doesn't count towards earnings, and you don't take an equity hit.

So the main question is, will SIVB suffer a bank run? Definitely the press is bad, but having sold their AFS portfolio, it means they'll be able to meet $25B of withdrawals without any issues.

However, they do have $170B of deposits. So if the situation is that people lose confidence in them, there could be some issues in meeting withdrawals above ~$35B. If there is a lender of last resort that can provide liquidity until the HTM portfolio matures, they'll survive.

If there isn't, and they have to take extraordinary measures such as selling the HTM portfolio at a loss, that would wipe out all of SIVB's $16.3B of excess equity. So ironically, today's situation is less a classical solvency and bad loans problem, but more a rate hikes problem

This issue is directly caused by the Fed and is the collateral damage from their rate policies. We saw a mini version of this panic from the UK pensions a few months ago. If this translates to bank contagion in the US, we are in for much more pain and definitely a recession.

Unfortunately, since banks are essentially a closed system, the deposits don't really go anywhere other than other banks. This means that we could see a wave of consolidation where smaller banks suffer a deposit drain instead of a systemic bank run.

// These are definitely volatile times. What are your thoughts/questions on this issue? For more insights and weekly discussion of markets, check out and subscribe to our newsletter!

https://www.pennyworks.com/articles


You can read the unrolled version of this thread here: https://typefully.com/crlfq/fvOwgJN

Silicon Valley Bank And the Modern Day Balance Sheet

I was a Senior Portfolio Manager in Bank of America's CIO where we managed over $500BN in US fixed-income assets.

Here's why I dropped everything to go into crypto and DeFi 🧵 👇 

2/ Blockchains disintermediate from the center: I traded with all of the major banks, and their "masters of the universe" actually get paid to do tons of rote tasks, they just rely on esoteric sheets no one else can use. Why pay 6 or 7 figs when you can pay a few dollars/cents?

3/ This leads to the second issue: Besides market risk, the biggest risk is actually operational. Many of these markets are highly specialized, the senior trader of TIPS going on vacation with some of his peers at other banks could lead to dramatic declines in liquidity.

4/ And that's not all, the clearing and settlement process is extremely painful. Many traders stay up late to recon what they think they traded vs the system thinks they traded. With blockchains, trade and settlement are one and the same thing. It becomes a non-problem.

5/ This matters cuz it's not enough just to do the same things slightly better, we need to do 10x better, because the truth is, the top 10% of the U.S. owns 77% of US wealth, it's clear as day that the capital markets are not working for "everyone", just the very few.

6/ But with DeFi and L2s, someone can begin saving and *earning* interest regardless of how little they have, whether they can maintain a minimum balance, whether they have a physical address or not, and whether they are "documented" or not. That's true equality of opportunity.

7/ Crypto is also more resilient than traditional finance: Even the recent downturn in crypto has given me more confidence on DeFi than ever before. When was last time any asset class deleveraged by $2 Trillion dollars within a few weeks with no major financial stability issues?

8/ No one stepped in to become lender of last resort. It was messy, each protocol made adjustments, and a lot of people were caught flat footed and suffered losses. But what's important is that DeFi continued to function as designed, which is more than I can say for CeFi Crypto.

9/ Even the fallout is more transparent. You know where you stand in DeFi at any point in time. But for CeFi, who knows when you'll get any recoveries from a bankruptcy process. Speedy in TradFi is measured in months and years.

10/ Crypto also unlocks the doors to unbounded innovation: Even NFTs, which had a wild ride up and down recently, are here to stay. There are now innovative financing products built on top of NFTs that span anything from a security, to a timeshare, to a membership pass, and so on

11/ Investment banking, the poster child of a business that relies on relationships, derives most of their value from connecting sellers of illiquid businesses with buyers. A legal and cheap way to financialize/securitize SMBs would unlock an inordinate amount of wealth.


12/ There are ~20,000 public securities on US exchanges, even including OTC securities (penny stocks). It took the US hundreds of years to get here. I see over 20k tokens on CMC right now. 

13/ There's definitely a quality difference, but my question is more fundamental. What does a world look like when there's billions of securities? Is it a richer world? Or would we bemoan the good old days? Would your answer change if we replaced "securities" with "cellphone"?

14/ Having a device that can access the entirety of the world's accumulated knowledge instantly is a game changer. Financialization can have a similar impact. Why? Cuz the world is becoming more intangible, and the most valuable assets are now  produced by minds, not mines.

15/ It could unleash a wave of innovation as never seem before as people are able to fund other people's ideas in unimaginable ways, giving everyone the opportunity to come up with the next big thing, not just those fortunate enough to have the right "profile" to raise capital.

16/ This could also give people the incentive to invest in education at a global scale, so that quality education would no longer be at the whim of your local government.

17/ There are too many potential use cases to think of, there's only one time I've ever felt like this before, and that's in the late 90's with the advent of the internet. It's funny that I'm only putting into words now, thoughts that have been with me for nearly a decade. 

18/18 But just like in the late 90's, there are more doubters than ever. So I'm here to say: Crypto will eat the world, and being in the passenger seat is just not enough, so strap on your seatbelts cuz we're going on a ride!

My thoughts on today's FOMC meeting: 🧵 

Fed Fund Futures are pricing in peak fed funds rates around 3.25%-3.5% around March of ‘23. Which is 1.75% higher than today and would imply the current pace of hikes can only be sustained for another 2-3 meetings before the Fed pauses.

Given the eye popping June CPI of 9.1%, the market is leaning towards a 75 bps hike this meeting. However, many agricultural/commodity/energy prices have come down dramatically since the highs a couple months ago. 

Anything less than 75 bps would be considered dovish and would likely cause crypto as well as equity markets to rally. If the Fed hikes 75 bps, people will have to read between the lines on what their expectations are for the next few months. There are 2 possibilities:

1.  Any mention of international cooperation will be seen as dovish since Japan and Europe are having a hard time matching the Fed pace given their structurally high debt loads. Cooperation inherently means either a slower pace of hikes, or some sort of intervention via FX swaps.

While FX swaps do not directly impact interest rates, the move will be perceived as accommodative, and markets would rally on the news.

2.  If they do not, given how consistent the Fed has been in communicating their vigilance against inflation, I can’t imagine there being any more hawkish news that could surprise the market. In that case, it’s likely that the post hike market reaction is relatively muted.

How do you guys think it'll play out? And let me know if you guys like these blurbs, happy to write more of these if there's interest!

Pre FOMC views 2022-07-27

TLDR on $LUNA & $UST:
$UST will slowly repeg, my guess is within 2 weeks. $LUNA will be minted to make that happen, meaning LUNA will underperform relative to rest of crypto, but will not go to zero.

What Happened though? 🧵 👇 

1/ On May 7th, $LUNA began dramatically underperforming relative to the rest of the market, down approx 8%-10% hitting low $60s. While that was happening, the $UST peg started to break. On curve, UST was sold as 4pool volume spiked to 500mm/day, sending the peg to mid/low 0.99xx

2/ Do Kwon came out to calm the markets, and there was some temporary relief later that day, where the peg (inversed ratio in the tweet) came back to around 0.995
https://twitter.com/muggle_magic_/status/1523077715070717954?s=20&t=TGkdA4RI-TKPxpSAr6Cr9A

3/ Some say it was a coordinated attack, I prefer to think of markets simply righting themselves over time, you could see a similar dynamic in January when reserves ran low. The pace was not as frantic, but the same dynamic applies.

4/ Once the snowball started rolling, my hunch is that those that were levered, i.e. used BTC/ETH to borrow USDx, swap to UST to generate yield had to de-risk lest they risk getting liquidated, so less of an attack and more like a rush to the exits. Hence the spiral down.

5/ Given this started on a weekend, it was also difficult for anyone to add liquidity into the system from fiat, so the only option was to keep selling crypto and delevering. 

6/ Meanwhile, on Sunday, @LFG_org enacts a plan to use $BTC reserves to stabilize UST.
https://twitter.com/LFG_org/status/1523512196965167104 https://twitter.com/LFG_org/status/1523512196965167104

6/ Once Monday came around, that's when all those investors exposed to crypto through ETFs and traditional futures markets were able to hedge, which essentially accelerated the selloff. Given the magnitude of the move, it felt like a instance of crypto shock bleeding into TradFi.

7/ Just from an size perspective, at the beginning of Monday there was still $12B just in Anchor, so it simply came down to not enough capital to support the selling pressure, the $LUNA market cap dropped below $UST, causing $UST to break from the peg violently.

8/ Monday's $UST volume: ~$5B traded on the centralized exchanges, and probably a few more on DEXs. The silver lining is that a huge amount traded at a discount, which reduced pressure on $LUNA since the only other way out was through the mint and sell mechanism which is capped.

9/ The deleveraging is continuing today, (TVL on Anchor down to $8B, $UST mcap $16B,  $LUNA mcap $11B, there's probably only $3-5B left to delever. Those that put into anchor late last year are breakeven despite $UST depeg to $0.90. LFG's assets of ~$3-5B is a floor for $LUNA

10/ Since the $LUNA mint process is rate limited, it will take a while for UST to regain the peg, but I think that'll work as intended. My hunch is that $LUNA mcap remains under pressure, with a low of $5-10B in the next 2 weeks, but unclear at what price given the mint process.

11/ For those that ask if $LUNA $UST will go to zero, my answer is no. There were some mistakes made to incentivize accelerated adoption, like using steroids to get short term gains, but is an algostable coin not valuable or impossible? Clearly $DAI exists.

12/ The better question to ask is, how many will there be? Likely less than a handful, and their worth will only truly be appreciated once they develop a track record of stability and sustainability. That doesn't just come down to mechanism design, but also prudent governance.

13/13 If you liked this tweet, please follow and check out my other threads!

https://twitter.com/crlfq/status/1445197638132019201?s=20&t=TGkdA4RI-TKPxpSAr6Cr9A

You can read the unrolled version of this thread here: https://typefully.com/crlfq/E4j4vrt

UST Depegs

For those deciding between Roth vs Traditional IRA/401k for their retirement, I've made a toy model based on future expected incomes, contribution rates, taxes and Monte-Carlo simulations of stock returns. Here's what I found:🧵 

2/ Effect 1: In high income scenarios, Traditional comes out ahead because you pay marginal tax rates for Roth contributions, whereas if you defer tax with Traditional and withdraw at retirement, it's taxed at standard income rates, which usually work out to be lower.

3/ Effect 2: The higher your portfolio's growth prospects, the higher the benefit from a Roth, this is because it increases your future retirement income, closing the gap between the above tax rates.

4/ Effect 3: he more you can save, the more value there is in a Roth account at retirement. So if you can do a Mega Backdoor Roth 401k conversion, you should.

5/ The analysis was done with real rates, i.e. adjusted for inflation. If you believe inflation is going to be higher in the future, then it'll be equivalent to lower real portfolio growth, and less value for a Roth account.

6/6 Lastly, if you believe tax rates are going up because brackets won't be adjusted for inflation AND the gov't will hike taxes, then Roth looks more attractive.

If you found this interesting, or are interested in learning how to code this up yourself, happy to chat!

You can read the unrolled version of this thread here: https://typefully.com/crlfq/JPAXQiQD8bHM