March 2023 Commentary

Cryptoassets are an asset class that famously made many rich but when asked, very few actually got rich. When you look at its return profile, with all of its spikiness and volatility, it's easy to see why. Bitcoin, an asset that is a 300-bagger over the last decade but with 90% drawdowns, has easily been declared deceased several times over. But aside from the price rollercoaster the reason why people have always missed the boat on cryptoassets is because they lacked a fundamental belief or thesis on "Why Crypto?" - their investments have always followed the news cycle which lags the price. Given the negative crypto headlines in 2022 and the March bank convulsions, did you know that bitcoin has appreciated over 70% in the first quarter, besting all the major asset classes most investors would allocate to?

Source: NYDIG, “1Q23 Review and Look Ahead”

This is a monthly commentary, so we’ll focus our attention on the month. March was truly an odd month in that two investors with very different mindsets would have had diametrically different experiences. On the surface, a macro-minded investor with a longer horizon would’ve seen that risk assets performed well with the Nasdaq Index posting a 6.9% gain and the broad digital assets market rallying over 12%. Three-month US Treasury yields eased slightly to 4.6% while on the long end, the 10-year eased to about a 3.5%, with the inversion still foretelling that a weakening economy is in the works. Nothing new there. Consumers continued to see relief from US CPI measures which dipped to a 6.0% year-over-year increase in February. By many counts, this was a good month for the risk assets.

The story beneath the surface was a different one. The banking industry convulsed as depositors, who had been leaving in droves over the past several months, caused enough panic to force the FDIC to take receivership of Silvergate, Silicon Valley, and Signature banks, and the US Federal Reserve to line up major banks to inject deposits into First Republic. The Fed also set up its emergency Bank Term Funding Program to provide banks with short-term loans on assets on their balance sheets on an as-needed basis, without resorting to asset sales. Three weeks later and, well, it’s been quiet. Risk asset prices have rebounded, bitcoin is back to the $30k level as of the time of this writing, and Ethereum maxis are looking forward to a successful Shapella upgrade.

The long-horizon investor, adopting a style of buying and patiently holding on to their investments, may be forgiven for missing out on the March madness. They would be left puzzled by the prognostications coming from the doomsayers such as Balaji or Hayes, calling for the unraveling of the fiat system and a $1M price on bitcoin in a period as short as 90 days. A rational investor, however, needs to adopt mindsets and tools from both investor archetypes.

Correlations – What they do and don’t tell us

Over the course of 2022, risk assets in general have been dominated by macro narratives from the war on Ukraine and disrupted supply chains to rising inflation and rising interest rates. Digital assets have been beset by their own “macro” events owing to some spectacular failures and fraud and the tightening regulatory oversight that followed in response. The Merge, a significant milestone reached in September 2022 on the Ethereum roadmap, was met with a shrug and a brief sell-off.

Correlations are an oft-used metric among investors, economists, and journalists to explain things big and small. The metric ranges from -1 to +1, and it measures the degree of how likely two series are to move together although in and of itself, the measure doesn’t tell us about the magnitude of the movements and certainly not the causality.

The extent of how much the macro events dominated investors’ mindset and the market behaviour can be observed from looking at the cross-sectional correlations of the top 50 digital assets by market cap over the past 3 years.

The correlations are measured over a 12-month window using monthly returns. The actual numbers are less interesting than understanding the trend of the intensity of the metric over the past 3 years ending March 2023. The blues represent lower positive to negative values, whites values around +0.4 to 0.5, reds above +0.6, and dark reds close to 1. The most recent 12 months exhibited a digital assets market that moved closely in tandem with each other as investors focused on global and digital assets macro issues.

Macro is so yesterday

Relying too much on correlations is a tricky endeavor. As a metric, the measure is subject to several parameters that need to be tuned, updated, and interpreted correctly to be useful.

For example, consider the correlation between bitcoin and technology stocks that are often cited for allocation a portion of one’s investments into bitcoin. Using the iShares US Technology ETF (IYW) as a proxy for technology stocks, below are plots of the correlation of these two assets using only slightly different parameters.

Both lines measure the correlation between the two daily return time series except one uses a 365-day observation window (black line) while the other uses a 90-day window. What seemed like a gradual increase in correlation between bitcoin and technology stocks over the last two years papers over a large variability in the behaviour of these two assets.

Which is correct? Neither and both. It depends on what questions you are trying to answer. For a long-term investor trying to understand whether bitcoin offers a differentiated long-term risk premium from that provided by technology investments, one should not be even looking at daily returns and should at least be looking at monthly returns. For the investor looking to market-time the short-term shifts in investor sentiments, a metric that samples the data at a high frequency and aggregates these samples over a shorter time window may be more appropriate.

For us to truly understand and forecast market behaviour, we would need to know the incentives driving the individual agents who trade and hold their positions. This is not directly observable, and thus, we default to statistical metrics for which we need to tune the measurement parameters. Done correctly, it can offer valuable insights into the markets, our asset selection and portfolio construction processes. We’ll be discussing more about this in an upcoming article.

While You Weren’t Paying Attention, Crypto Beat Your Portfolio

One Year!

Complexity and Unknown Unknowns