May 25, 2022
When it all comes down
There'll be nothing left to catch you but ground
It's calling your name and filling your head
With delusions of glory
Is that how you're gonna write your story?
Down in your time as a high-flying flame out?
Sucking on what's left of your trust fund?
Sucking on the end of a shot gun?
But there's more here than meets the eye
The real story is under the surface
We're all so in love with the artifice
We don't dare look too close
“A Blessing and a Curse” – Drive-by Truckers
I am not crypto-native. I come from the TradFi world having spent 30 years in traditional financial services, with 10 years in investment banking and 20 years in a hedge fund. However, I have followed the crypto space quite closely since 2017. That year, I saw a large number of former colleagues getting involved in Bitcoin and other digital assets projects. I saw the rise and fall of Bitcoin. I watched the crypto winter from a distance, telling myself I would get involved when Bitcoin got to 5000. Why? For me the decision was twofold: it was a good technical level, and the amount of time it took to get there had allowed a sufficient period of grieving. See, even though I spent my career as a fundamental analyst, and now teach the same ideas at university, I also know the behavioral aspects of the market are very powerful. I had seen it in every asset class and in every time zone around the world. It was not unique to any particular type of investor or investment. It was part of human nature.
In late 2020 and early 2021, I listened with great interest as the bull cry was that institutions were getting involved. It really started with MassMutual buying $100bb in Bitcoin in December 2020 but 2021 saw more and more institutional investors get involved. The bullishness was not just about the validation of a product, it was about the major flows that were coming in. “Did you hear Paul Tudor Jones says to buy Bitcoin?” Heard that all along. As I heard it, I thought to myself, people better be careful. Institutional money is a blessing and a curse. The flows are great on the way up, but they can be painful on the way down.
Why do I say this? Is this because institutional money has weaker hands than other investors? Not at all. I say this because having worked with scores of asset owners, I understood what drove their asset allocation decisions. Having worked at multi-strategy hedge funds, I understood how money was allotted to the different strategies or ‘pods’ within the hedge fund. Watching all of this, I understood that the more institutional money got involved in crypto, the more crypto would look like other markets and not the other way around.
For review, let’s look at the correlation of Bitcoin to other assets. I have taken the chart below from bitcoinist.com and it shows the rolling 30-day correlation of Bitcoin to the S&P, Nasdaq, US Dollar Index and Gold. What is the message? Bitcoin and risky assets are getting more, and not less, correlated:
Why is this the case? That is always the ultimate question i.e. what is correlation and what is causation. Let me posit one possible theory. Hedge funds and proprietary trading firms are likely the most active institutional investors involved in all asset classes, and not just equities and crypto. These firms largely use a model called Value at Risk when deciding how to allocate the investor or partner money across various strategies or ideas. According to Wikipedia: “Value at risk (VaR) is a measure of the risk of loss for investments. It estimates how much a portfolio might lose (with a given probability), given normal market conditions, in a set time period such as a day.” VaR can be defined as the maximum loss an allocator is willing to risk over a certain time-period in a certain strategy.
For example, the CIO of a hedge fund will likely not tell you “Here is $250mm of investor capital to manage” but more likely say “You have $10mm of 99% 10-day VaR”. What does this mean? If a portfolio of stocks has a 10-day 99% VaR of $10 million, that means that there is a 1% probability that the portfolio will fall in value by more than $10 million over a 10-day period if there is no trading. This is what the CIO is willing to lose on you in ‘normal’ times. The CIO is doing the same for other portfolio managers/strategies/pods in the firm, hoping she has many uncorrelated streams of alpha. The idea is even if one blows up, the others can be there to make money. In theory.
One of the key metrics used in calculating VaR is the implied volatility (sometimes realized volatility) of the asset. This means that as volatility goes lower, the perceived risk of the strategy goes lower. Thus, the PM has more risk to take and is often strongly encouraged to take it. As all PMs are doing the same thing, the same strategies that were making money continue to press the same ideas. This leads to lower volatility and therefore lower VaR. Virtuous circle. However, the converse is also true. When volatility moves higher, the VaR for the same portfolio is now higher. These means strategies must de-risk. This can often lead to higher volatility and now a vicious cycle. When there is an exogenous shock (Covid), volatilities spike higher even without the de-risk but that volatility becomes a catalyst for the first wave of de-risking until something happens that dampens the volatility. This is why many levered investors are looking for that Fed put. It isn’t the flow into their asset as much as the dampening of volatility.
So, in a crisis, the correlation of assets and the correlation of strategies goes toward one. This means those uncorrelated streams of alpha all of a sudden are correlated. This means the firm is not as comfortable as it was before. This means those expected losses can be larger.
As more firms are investing in cryptocurrency as well as other assets, losses in stock strategies can lead to reductions of risk in cryptocurrency strategies. A tech long/short strategy blow-up may mean the firm VaR is now higher so other strategies have to reduce. Similarly, a shock and losses in a yield-farming strategy can lead to a higher firm VaR and the need to de-risk in credit and convertible bonds. Firms that are investing in all assets, will see and then drive the correlations of these asset classes closer together and not further apart.
You might be saying, “But these are traders, or fast money, strategies. This is not how the end-investors react.” You would be right. The asset owners – high net worth families, pension funds, endowments, life insurance companies – do not allocate based on Value at Risk. However, these asset owners use Modern Portfolio Theory which relies on the Markowitz Efficient Frontier. They probably improve on this with Black-Litterman allocation models.
Markowitz created a formula that allows an investor to mathematically trade off risk tolerance and reward expectations, resulting in the ideal portfolio. Modern Portfolio Theory works under the assumption that investors are risk-averse, preferring a portfolio with less risk for a given level of return. These ideas are the cornerstone of finance and investing taught in business schools. I know, I teach in a business school. The Black-Litterman approach tries to improve on this by using an investors expectation of return, risk (volatility) and correlation going forward instead of trailing measures. The past is not indicative of future results after all.
Let’s go back to those earlier correlation charts. This time from Barron’s. Asset owners were loving crypto assets. Yes, volatility was high. However, returns were strongly positive and, most importantly, returns were uncorrelated if not negative correlated to other asset classes (2019). This meant even with high volatility of the asset class, adding in investments could reduce overall portfolio risk because it was negatively correlated. Much like that hedge fund CIO, the goal is to find as many uncorrelated (ideally negatively correlated) streams of alpha as one can find. Bitcoin was it.
I might add, DeFi was considered even better. Why? DeFi applications were much easier to understand for TradFi investors, because large numbers of them were just similar ideas on a new market. These were disruptions of a business that they understood because it was disrupting THEIR business. Whereas Bitcoin may seem more theoretical to many, DeFi was more straightforward. Thus, institutional money came in. Asset owners went to hedge funds to gain access. Hedge funds embraced the flows and put it to work. We had the makings for the rally of 2021.
What caused the change in sentiment in 2022?
It is hard to tell. I would posit that it is largely driven by the change in discount rates as a result of changing Fed policy. This is driven by inflation. Higher discount rates means high growth equities and credit, the big winners in 2021, were now at risk. These strategies started to suffer meaningfully leading to losses in other strategies. The war in Europe, the Chinese lockdown and the collapse of a stablecoin are not helping the situation because each and every one of them leads to higher volatility. What does higher volatility mean? It means higher VaR which means risk-reduction which means higher volatility. You get the joke.
How does this all end?
It is said that volatility is like a forest fire. You never know what the spark will be that starts it, but you know when the conditions are ripe for a fire – drought conditions, poor forest management etc. What does that look like across asset classes? Drought is the same as the absence of liquidity. Central banks are removing liquidity from the global economy. Sanctions have removed other sources of liquidity. Poor forest management is like misallocated capital. How does a forest fire end? We try to contain it if we can. We ultimately may need to douse it with a lot of liquidity. The latter does not seem like it will come from central banks just yet. We have seen 3 hikes from the Fed and there are still 9 more priced in until March of next year. If the expectation comes lower, this is help at the margin. Other central banks in China and Japan are starting to add liquidity. A last source would be investors that find the potential forward returns too good to pass up. Sometimes this can end quickly and there is a V-shaped recovery as we saw in Covid. Other times it is more drawn out. However, in order to put on the volatility fire, we need liquidity.
More importantly, how do we manage this risk in our portfolios?
This source of risk comes because volatility is moving higher. If we want to protect against potentially higher or persistently high volatility, we want to use options in the expression of our views. Adding in this convexity to long or short ideas makes are portfolio ‘anti-fragile’ as Nassim Taleb would say. Not only does it not get hurt by volatile events, it thrives from volatile events. Long option ideas are not just insurance, but can be a tactically offensive approach to optimize the implementation of ideas.
Others may feel that the worst is being priced into markets. Options are not only insurance, but also yield or income enhancement. Even thought rates are heading higher in relative terms, in absolute terms they are still low. Investors will still be desperate for yield. Cash-secured puts on coins you really like, or overwrites on coins you already own are a way to take advantage of the higher price of volatility and generate some income until capital gains return. Surely there will be structured note issuance that will have embedded option selling involved.
Paradigm can help facilitate either of those approaches to working through the current market environment with its extensive liquidity network of 800+ counterparties.
Right now it is hard to see through to the other side, but this current market tumult will end. We will come out on the other side and we won’t be thinking about this asset class or that asset class. We wont be thinking of crypto as a whole, but will start analysing coins by their various fundamental value proposition (i.e. whether it be a store of value, fuel for smart contracts or perhaps the equivalent of an equity stake in a crypto project etc). We will be thinking about how to fund that idea and what its potential market capitalization could be. Correlations do not stay near one forever so we will begin differentiating across ideas.
It also doesn’t mean that institutional money is leaving the digital asset world. It is most likely not only here to stay but here to grow over time. How much it grows will depend on future expected returns, volatility and correlation. These expectations will drive the inputs to the Black-Litterman allocation models and determine how much to allocate to each asset class. High Sharpe ratios and low correlations will continue to be preferred.
What might change in the crypto thought process though? While there was once the thought that the flow of institutional money was a blessing, we are finding out that it is a blessing and a curse. Or as the Drive-By truckers sang:
It's a blessing and a curse
Watch out, Eugene, you don't make things worse
Wild dreams come true, what to do then?
Confusion and glory