News & Resources

Digital Dives: Risk is a Four-letter Word - Vol. 6

March 4, 2022

Risk tends to carry profane connotations, but it isn’t necessarily something to shy away from. In fact, many of humanity’s greatest achievements have been attained by overcoming hazard. Pushing boundaries can create new high watermarks against which we might measure future successes. Therefore, I think we’d be well-served to reframe the common perception of risk as something to be avoided and consider it as the fee paid for earning a return. However, simply taking this perspective doesn’t make it any easier to navigate unknown waters. Sometimes we’re “paying fees” that go unnoticed, until it’s too late. Humans are hard-wired to be loss averse, which is the facet of risk that we’re ultimately trying to contain. To do so, we must actively train to set ourselves up for success – these efforts are what I call risk management.


TextDescription automatically generated


The History of Risk

In Western civilization, the first published work in the field of probability theory came from a pair of French mathematicians towards the end of the Renaissance. In 1654, Blaise Pascal and Pierre de Fermat exchanged letters in which they discussed how to solve a particular gambling problem. Prior thereto, the concept of randomness wasn’t something the average Westerner thought about because all happenings were pre-ordained by some omniscient being(s). This implies that the concept of risk is less than 500 years old. That we struggle to conceptualize and quantify hazard isn’t so unbelievable in this context. 

About a century after the first report on probability calculus, Thomas Bayes leveraged statistical methods to illustrate how using math to blend new information into existing data can help us make better decisions and a hundred years later, Francis Galton discovered mean reversion. In 1952, Harry Markowitz expanded the work of these early scientists to the realm of portfolio management with his Efficient Frontier and Modern Portfolio Theory (MPT). This set capital allocation on a course to uncover investments that produce the highest possible gain for a given level of risk. The standard deviation of returns was assigned as the metric to measure the latter. The simplifying assumptions in the model have been criticized over the years, notably that investors act rationally. The idea that volatile assets are only desirable if they produce a commensurate level of return is intuitive, but humans tend to be more negatively disposed to potential losses than gains. A sensible investor would understand that an asset whose price cascades should have the capacity for similar upside moves if it’s been well-selected, but in practice the experience is more like this:

DiagramDescription automatically generated

https://thedecisionlab.com/biases/loss-aversion/ 

A Double-edged Sword

Investors would never balk at a big gain, but they should be mindful that such a move could be expected on the downside too. That said, if we’re in the business of seeking out returns, then our work should include an attempt to understand the potential sources of peril in our strategy. Capital is the lifeblood of modern society and often it’s being allocated with the principle that volatility = risk, but specifically, it’s the downside moves that are seen as threatening. While the ways an investment thesis can fail are innumerable, there are two broad classes of permanent capital impairment:

  • The value of your stake falls and never recuperates due to some underlying fundamental deviation, or
  • You sell your holding for less than you purchased it, and this produces a loss

Bad luck or poor analysis can result in the former, but the latter requires the additional ingredient of individual action. If your investment thesis has changed due to the revelation of new and adverse information, then exiting a holding can be prudent. Uncovering a different asset that is more deserving of your capital is another good reason to make a change as well. As Howard Marks notes: “There certainly are good reasons for selling, but they have nothing to do with the fear of making mistakes, experiencing regret and looking bad.” We must strive to control our emotions. 

In the context of investing, there are two tools that can help avoid capital impairment. The first is diversification and its benefit was beautifully detailed in Mr. Markowitz’s work. AQR has written a few times on the topic, including this gem, which outlines the benefits in an adverse situation:


Chart, line chartDescription automatically generated


The orange and dark blue lines above represent the worst-case outcomes of 100k trials for stocks and bonds respectively. The dotted light blue line is the 60/40 weighted average of those outcomes, while the solid light blue line incorporates the benefit of diversification “because worst outcomes tend to happen at different times for different assets — and this can become economically meaningful at multi-year horizons”. The second loss minimization mechanism is a profound understanding of the assets being acquired because this knowledge can:

  1. Help avoid duds; and 
  2. Fuel conviction to maintain a forward-looking perspective during periods of adverse volatility. 


Are Digital Assets Risky?

Messari contemplated whether crypto’s ascension to mainstream finance has shifted the narrative around crypto as a risky asset within institutional portfolios. Tokens continue to demonstrate substantial price volatility, but so far, they’ve also delivered massive gains. Because these cryptoassets have relatively low correlations with traditional holdings, the portfolio math makes them remunerative and risk-enhancing. During periods of extreme market stress, risky assets tend to move in unison, but crypto has demonstrated some idiosyncrasies during the most recent rout. As Tomasz Tunguz notes:


TextDescription automatically generated


Web3 assets certainly have speculative attributes. That said, risk management is often less about volatility than it is about avoiding capital impairment. In traditional finance, the insurance industry is built on the idea that by aggregating risk, evaluating sources of hazard, and applying probability theory, shrewd analysts can profit from the apprehension of others. Insurance companies are some of the oldest out there and the industry is over $5T in size. 


How do we manage risk in this new frontier? 

As investors, understanding the digital assets we intend to hold is of utmost importance. This will help us avoid making investments whose prospects are unlikely to pan out and will give us the conviction to hold steady through violent declines. If these aren’t core competencies, then it is prudent to seek out a trusted advisor who can shepherd your capital.

It helps to have a framework when managing risk in traditional asset markets. The Global Association of Risk Professionals (GARP) maintains a curriculum for would-be managers of financial hazard and they classify risks according to three broad pillars:


DiagramDescription automatically generated

https://slideplayer.com/slide/12993085/ 

Let’s contemplate each in a digital asset context:

Market Risk relates to the potential for losses due to factors that affect the performance of investments in the financial markets broadly. Also called systematic risk, this source of price action can’t be diversified away:

  • This is where the bulk of crypto-related academic research has focused to-date, which makes sense, since the datasets are robust.
  • At the onset of the global pandemic, we witnessed a classic case of market risk. It was a period of extreme volatility and tradeable assets were sold indiscriminately.
  • I’m over-generalizing here, but market risk is relatively easy to contain because hedging instruments are readily available and dramatic moves tend to be smoothed over longer time periods. However, insulating beta can be expensive and in a system funded with leverage, liquidity squeezes and forced selling can lead to significant pain.

TextDescription automatically generated with medium confidence

https://twitter.com/jackbutcher/status/1485609022799716353

Credit Risk is the possibility of a loss resulting from a borrower's failure to repay a loan or meet contractual obligations. Ex ante analysis is difficult, but a careful assessment can help minimize shortfalls. Credit is commonly assessed along two dimensions: Probability of Default (PD) and Loss Given Default (LGD):

  • Consumer credit risk can be measured by the Five Cs: credit history, capacity to repay, capital, the loan's conditions, and associated collateral.
  • Digital lending is relatively new, so reliable historical data is limited and makes estimating default probabilities more difficult, but that’s starting to change. Currently, most decentralized loans are over-collateralized, which helps protect lenders (reduces LGD). However, this might be too punitive for borrowers. Developments here will be something to watch.
  • Below, you can see that there have been episodes of extensive liquidations in DeFi, but the system has remained intact.
Chart, bubble chartDescription automatically generated


Operational Risk arises from the uncertainties and hazards a company faces as it conducts its day-to-day business activities. Sometimes called human risk, it results from shortcomings from within the organization like procedural or control weaknesses: 

Some crypto-specific Issues

To supplement the framework above, it’s important to consider some nuances related to web3 assets. Moody’s is a stalwart of financial risk analytics and they recently published a note in collaboration with Gauntlet titled Block by Block: Assessing Risk in Decentralized Finance. Lily Francus co-authored the piece and she remarks that: “As DeFi has gained the attention of traditional financial institutions globally, it is likely we will see continued convergence due to the high complementarity between traditional financial services and DeFi”. I’d recommend working through the whole piece, but here’s a summary of their findings:


Graphical user interface, textDescription automatically generated


Note that the first three items above are systemic in nature and as such they can’t be diversified away. However, the impact of problems related to the other issues, can be reduced through correlation management and owning a basket of tokens. 

Regulations are a hot topic in web3 and while the hazard posed by adverse legal decisions could have profound effects, the industry has made significant strides in its ability to influence the outcome of the inevitable oversight that is forthcoming. Andreessen Horowitz formalized a vision for policy response in “How to Win the Future” and the VC firm continues to engage with officials to keep the US’ innovative ecosystem intactAndrew Yang launched Lobby3 to further the cause as well. He spoke at the ETHDenver conference, which you can view here (the good stuff kicks off around the 1:18 mark). While managing this risk still requires a probabilistic approach in considering the relative impacts/chances of different scenarios, it’s cool to see advocates trying to tilt the odds towards a favorable result. 


Conceptualizing the different risks related to digital asset ownership should involve a blend of quantitative and subjective insights. The problem with relying too much on historical data when modeling different scenarios is that we simply don’t know what the future could look like. As Nassim Taleb puts it in Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets (a favourite book of mine): “If the past, by bringing surprises, did not resemble the past previous to it (what I call the past's past), then why should our future resemble our current past?” We should inform our risk management with historical time series, but this approach has its limits. Alla Gil explains why advanced strategic asset allocation techniques are made more robust by modelling unprecedented outcomes. This makes risk management more of an art than a science and as Peter Bernstein points out in Against the Gods: The Remarkable Story of Risk (another one of my favs): 


TextDescription automatically generated
TextDescription automatically generated


Arca is an institutional investment manager and thought leader in the web3 space. Their Digital Assets Fund lost 34% in January but was up 315% in 2021. This is a good example of crypto’s volatility in action and their most recent commentary provides a real-life demonstration of the concepts covered above. The investor letter elaborates on some of the systemic and idiosyncratic elements that worked against the portfolio, a summary of various risk metrics, an admission of anticipated scenarios gone awry and an industry outlook:

TextDescription automatically generated


Remember, risk is the price we pay to earn a return. Our evolution keeps us focused on the downside but daring against the odds can lead to fantastic results when executed mindfully. It’s essential to stay humble and open-minded when contemplating both sides of the return expectations curve.

After all that, do you think digital asset ownership is worth the price of admission?