November 24, 2022
"In economics, things take longer to happen than you think they will, and then they happen faster than you thought they could.”
The last 10 days have rocked the crypto community and the aftershocks of the FTX Chapter 11 are still being uncovered. Sadly this affects a significant number of investors both institutional and retail and we have yet to fully understand what caused this – was it just too much leverage, misallocation of funds into risky investments or straight up fraud?
Without knowing any details, I can tell you the scope for traditional finance players to invest into crypto has materially reduced. And to truly understand the TradFi mindset, we need to understand the history of quantitative easing and asset/liability matching, which has contributed to how investment decisions are made.
For the better part of a couple years, if not more than a decade, there has been a concern over the amount and magnitude of easy money in the financial system. With so much liquidity essentially at zero cost, the concern was capital being misallocated. It wasn’t a question of if, it was a question of when, this would all come collapsing upon itself. However, it didn’t. We all heard about TINA (There Is No Alternative), which took asset prices to new highs. How could this happen?
As Carmen Reinhart and Kenneth Rogoff told us in their book This Time is Different, in the immediate period post the Great Financial Crisis (GFC), the financial markets would witness a period of financial repression in which governments would implement regulations that force financial institutions to own more sovereign debt either as captive domestic audiences or heavy reserve requirements.
How did this play out?
In a few different ways:
• Under the Basel III regulation, post the GFC, banks were required to hold a larger percentage of high-quality liquid assets (HQLA) on balance sheets. These HQLA were deemed to be sovereign debt.
• Similarly, under the Solvency II regulation, insurance companies face a similar mandate where the risk-weighted asset charge for sovereign debt has zero and for assets such as corporate credit and equity were penalized. This particularly affects European and Asian companies.
• Pensions funds this century have been required to do asset-liability matching. With aging populations, this means more bonds must be bought.
What did this ultimately lead to? This buying by financial companies with the excess reserves created by the central banks lead to real rates moving lower across the curve. In fact, one of the principles of finance is the time value of money, whereby investors demand a higher rate of interest the longer they go to lock up their money. We can see that in the decade plus after the GFC, this term premium consistently fell from a peak of 3% to a low of -1%. Yes, that is right, investors were not only demanding a premium, but they were also willing to pay to lock up their money longer.
This drove real rates negative as well as would have been predicted by Reinhart and Rogoff. It wasn’t the first time we saw this response. In fact, they highlight how we had seen this in the immediate aftermath of World War II. We can see how the post-GFC and post-WWII periods both stand out from the “normal” in red below, in that the negative real rate tail was significant.
Ultimately this led to a bubble. However, it did not directly lead to a bubble in risk assets as most investors immediately pivot to. In fact, we had a bubble in duration. This bubble in duration impacted all assets. Most investors typically only think of duration in regard to fixed income. Duration is commonly understood to be a bond’s sensitivity to interest rates. More generally, duration is how long it will take for a bond investor to be paid back the bond price by the bond’s cash flows. Using this same definition, we can see that all investments have duration. For stocks, a defensive name that has a high dividend will have a short duration. A high-growth tech stock with negative earnings and where all value is from implied future growth will have a long (perhaps infinite) duration.
This duration bubble was caused not only by the combination of financial repression but also by central bank activity in the bond markets, particularly during COVID-19. You see, the initial QE from the Fed created reserves on banks’ balance sheets. It was thought the banks would both buy bonds and lend the money out. In fact, most banks took the money to repair the hole on their balance sheet. It only inefficiently made its way into the economy. Eventually the Fed and other central banks bought bonds directly. During COVID, the Fed and other central banks were much more direct about the balance sheet growth going directly into the market and the economy. This squeezed investors who also had to buy duration for regulatory reasons. Central banks were buying the same duration that financial institutions needed to buy. The result was a massive squeeze in bond markets, which peaked in late 2020 as the amount of negative-yielding bonds (these were negative nominal yields) globally peaked at $18 trillion. In fact, in 2020, the best performing security was the Austrian 100-year bond, which outperformed even FAANG stocks over this period.
This squeezed investors out of bonds and into other markets. As fixed income investors bought dividend stocks for the bond proxy, conservative equity investors bought growthier names. As growth investors got squeezed out of tech, they turned to cryptocurrencies. As cryptocurrency traders got squeezed, they bought Lamborghinis. Seriously though, we can see the bubble in duration ripple through the market in 2020 and into early 2021. In the graph below, the amount of negative yielding debt, the U.S. aggregate bond index, the Nasdaq stock index and Bitcoin. We can see the duration bubble get blown into 2021. What happened then? Central banks started to raise rates across the world. As bond investors know, when rates are going higher, one tries to reduce the duration in their portfolio. This unwind of duration rippled through all assets on the way down as it did on the way up, just more quickly.
As this duration bubble unwinds, it is impacting the volatility in the “macro” markets more than the “micro” markets. We can see this is the demand for hedging across the different markets. One of the big issues impacting many markets right now is that the risks are macro, and yet so many people are trying to make their investment decisions in a micro framework. The chart below shows the volatility of various assets. We can see the Treasury market volatility index (MOVE) and the FX market volatility index (JPM VXY). There has been a larger spike in these indices than in the VIX Index of U.S. stocks or in credit spreads (measured here as the Moody’s Baa Corporate spread less Treasuries). While these volatility measures historically move together through time, we have seen a very real disconnect between the macro and micro indices this year.
A big reason for this is that there is now more debt on government balance sheets than on either corporate or consumer balance sheets. One of the “solutions” post GFC was for governments to take on more debt as both consumers and corporations de-levered to get themselves healthier positions. We can see this in the chart below. Post the GFC, the purple area (Federal debt) has clearly grown while the green area (Households) has shrunk. Corporations had shrunk but with rates near zero, started to grow again, albeit slightly.
It is not surprising to see the stress in the US Treasury market this year when we understand that the vast majority of debt growth over the past decade has been on the government balance sheet more than corporate or consumer balance sheets. In fact, the U.S. debt to GDP ratio has more than doubled from just over 60% during the Great Financial Crisis to over 130% at this time.
So, what does this mean for crypto? It has major implications for all asset classes. The U.S. Treasury market is the anchor pricing market for other sovereign debt, corporate credit spreads, commodity and FX futures and forward, for equity multiples and discount rates on riskier venture investments. A higher rate and a higher level of risk in the funding market of choice for investors will have major ramifications for the types of projects that see funding. It has portfolio implications for how institutional investors may look to invest. It has implications for risk parity and liability-driven investing portfolios that were constructed on the premise of negative correlations between assets that are now positively correlated.
After all, it isn’t only crypto and the riskiest parts of the investment universe that is struggling this year. Pensions and endowments – the so-called safest parts of the investor universe – are also struggling under the weight of changes in market pricing. Bridgewater became the largest hedge fund in the world because of the creation of the risk parity concept. Put simply, it is the notion that assets in one’s portfolio should be weighted based on the contribution to risk and not in some fixed percentage (e.g. 60-40 portfolios). Fixed income was an anchor part of most of these portfolios because it was a lower volatility asset which was negatively correlated to risky assets in the post-GFC world. Thus the “all-weather” portfolio could be created, where bonds would do well when stocks struggled and when bonds struggled, stocks were doing well. This worked because the market was focused on growth and central bank easing, not on inflation and central bank hiking. 2022 has changed all of that. Risk parity has had one of the biggest drawdowns in the history of the strategy, even though the economy has yet to go into a recession.
However, risk parity is not the only portfolio relying on the negative correlation between bonds and other asset classes. A liability-driven investment or investing is primarily slanted toward gaining enough assets to cover all present and future liabilities. The amount retirement plans are expected to pay out to their members in the future are also known as liabilities, and so-called “liability-driven investing”, or LDI strategies, aim to match the value and time horizon of their current assets to those future liabilities
Liability-driven investing struggled in the U.K. because the pension market is £2 trillion. The U.K. Gilt market is also £2 trillion. Don’t forget about the banks and insurance companies that must also buy Gilts for regulatory reasons. There were not enough Gilts to go around. How does Wall Street respond? By creating synthetic Gilts through derivatives. This freed up cash for pension funds to invest in uncorrelated asset classes from stocks to EM debt to private equity and infrastructure. And it all worked well…until it didn’t.
While each pension had a “bespoke” solution, firms such as Blackrock were in the news for being the trigger on selling Gilts leading to BOE involvement. When the margin call came, pensions couldn’t liquidate other investments fast enough. You can see what happened to an LDI fund, for example.
You can see below how over the last 18 months, the correlation of equity indices (INDU, NDX) with bond benchmarks (LBUSTR, EMB, HYG) and crypto (XBY, XET) has been quite positive.
However, the pre-COVID years – when inflation was low and growth was the concern – saw negative correlation for U.S. bonds and close to a zero correlation for crypto. This meant these assets were diversifying for portfolios as investors look for as many uncorrelated streams of alpha as they can find.
All of this has major ramifications for crypto markets and crypto projects:
1. First, the market clearly has to get through the latest episode of fraud and hacking like we have seen in FTX. On the back of Luna, Celsius and 3AC, this has been enough to take crypto from “I need to have some investment in case I get any questions” to “I can’t have any investment so I don’t get questions” in the minds of many institutional asset allocators.
2. Once that is clear, allocators will be rebuilding their forecasts for return, volatility and correlation, the key inputs to a Black-Litterman model or Markowitz mean-variance frontier. Positive forward return expectations and falling expected correlations are key.
3. There will continue to be overhang in the duration market given the excess debt on government balance sheets and the wrong-footed positioning in LDI, risk parity or 60-40 portfolios. Projects that have a shorter duration (read positive cash flow) will be more likely to get funded.
4. Finally, a lot of money has come into crypto from the venture capital world. This is where the pain is. This will take time to work through as investments are written down, “down rounds” of capital are raised, and governance structures are improved. This is a positive long-term development but it will take time.
2022 has been a year of macro risks where correlations have risen toward one. The history of investment has a number of these episodes. They pass with time, with adjustments and with the loss of capital. However, in order to understand how it unfolds and where the opportunity is on the other side, it is important to know where the source of pain is and where the source of relief can come from.
Best of luck and stay liquid!