Risk Update – December 2020

Most of us are loosely familiar with what has been going on in the Venezuelan economy over recent years. Simply put, another in a long history of failed systems that tried to defend their position by debasing their currency. The below chart of USD/VES (the current revised version of the Venezuelan Bolivar) is what full-fledged hyper-inflation looks like. Few people would disagree with that statement.

Figure 1: USD/VES Spot Rate

Source: Bloomberg

Likewise, the below chart of Bitcoin versus a debasing USD might be considered early-stage hyper-inflation. At the very least inflation. We like to think of this chart as the report card for central bank competence (it is, of course, on an inverse scale). They should look at this every morning and consider how the world is judging their policies.

Figure 2: XBT/USD (Bitcoin)

Source: Bloomberg

These below price indices are not useful tools in measuring or describing broad inflation, maybe even deceptive tools. The US CPI index has grown at an annualized 2.24% (doubling roughly every 32 years) over the last 30 years. The PCE Deflator Core Index (the Fed’s chosen price stability indicator) has grown at a (supposedly) disastrously slow annualized rate of 1.82% (only doubling every 39.5 years) over the last 30 years. The Fed, as you are all aware, believes that prices should double, at a minimum, every 36 years (2% annual inflation) and that the basket and methodology behind the PCE Deflator is the mandated set of prices to determine the achievement of that objective. (Based on the last two years of 180% annualized growth rate of Bitcoin, it is doubling every 146 days).

Figure 3: US CPI and PCE Indices

Source: Bloomberg

In good old ‘Classical Economics’, macro-inflation is simply the growth of money and credit, which ought to make sense. Price indices are merely a measure of whether that particular slice of things is showing symptoms of inflation. An increased supply of money and credit is like an increased flow of water: it has to go somewhere. We can use something like M2 Money Supply as a better proxy for “true” inflation (still just a proxy as it won’t fully capture the growth of all money and credit). Figure 4, a long-dated chart of US M2 money supply, breaks into three regimes of growth/inflation. 1990-1995 M2 grew at an annualized rate of 1.88%, back in the day when you might say there actually was a mere 2% of inflation. 1995-2019 where it grew at circa 6.15% annualized. Finally, in the new world of exceptional monetary accommodation, 2020 where M2 expanded, aka inflation, by 24.69%.

Figure 4: US M2 Money Supply 1990-2020

Source: Bloomberg

It is at this point that some might ask the question; if the water has to go somewhere, why isn’t the circa 6% 30+ year annualized growth of M2 showing up in the CPI? Where is it going? A very simple visualization can help explain the obvious answer to that: asset prices. We show below as a good proxy for asset prices in general, the S&P Index, but of course houses, buildings, paintings, yachts, bitcoin, whatever would all be getting their share of water working through the cracks.

Figure 5: US M2 Money Supply, SPX Index, US CPI Index. Log format and normalized. 1990-2020

Source: Bloomberg

Some small amount of the water goes into consumer items captured in the basket tracked by CPI. A lot more goes into asset prices. Not surprisingly, those with a lower portion of their income going to consumption, tend to be nearer to the originating sources of the flow of water. Thus, the direct correlation of money and credit growth (inflation) to wealth segregation. To get a good idea of how devastating this is, we need look no further than the growth in median income of this same period.

Figure 6: US Household Median Income

Source: Bloomberg

Over the 30-year period shown above, median income has grown at a mere 0.65% annualized rate, only doubling every 110 years! Earnings aren’t even coming close to keeping up with the price indices, much less the debasement of money expansion and related asset inflation. Easy to see the wealth segregating implications of those that own assets versus those that earn wages. This is truly the Road to Serfdom, as detailed by Friedrich von Hayek. In essence, the only means to grow the capital to acquire some desired asset in the future is to own assets today. Income, and the saving of that income at effectively no yield, cannot keep pace with the debasement of fiat money against desired assets.

This, obviously, is not a great way to run and sustain a system. One can imagine the esteemed central bankers sitting around pondering the epic question of whether to allow the necessary risk and segregation reducing, but hugely painful, default/recession/deflation cycle to take place and correct some of the unsustainable imbalances, or to keep pursuing inflation as the only means of sustaining things a little bit longer, but at terminal risk to the entire financial infrastructure and social fabric over which they preside (unelected). As we like to say, no past reserve currency has so far survived forever. Sure, they will try to pander to their claims of economic science and the inevitable “this time is different” refrain but, to the extent they aren’t too far down the cognitive capture trap, they are indeed aware of this trade-off.

For an incredibly well-crafted explanation of these circumstances, we would refer you to Friedrich von Hayek’s Nobel Prize speech in 1974. It is a great piece, with plenty of bits that could have come straight from writings of Nassim Taleb or Per Bak.

https://www.nobelprize.org/prizes/economic-sciences/1974/hayek/lecture/

We can go on and on as to the risks to the system-as-we-know-it from endless debasement/inflation, the history books are full of it, but it is not our job to fix the world (though we are up for giving it a try if anybody wants to offer us the position). It is, however, our job to help people protect and grow capital. Given that we have no crystal ball that will tell us what is going to happen in the future, we stick with things that we can rely on, ie math, and, as you may be aware, math leads us to one key solution in improving compounded returns: convexity.

We think the answer is not in guessing, or timing, the ongoing success of debasing currencies to the benefit of inflating asset bubbles, nor the inevitable eventual busting of these bubbles. The answer is not about guessing whether inflation or deflation will win out, or when one will switch to the other. The answer, we believe, is in constructing a portfolio that is resilient and optimizes compounding regardless. The answer is not to tie up massive chunks of capital in things that neither benefit from or participate in the inflation/credit creation, nor protect against a subsequent bust and sharp period of deflation/credit destruction. More directly, we would urge investors to stop holding near zero yielding cash, fixed income and credits; stop paying people to run levered carry trades at all time lows in yields, spreads, and vols. There may be times when there are good trading opportunities to be “long bonds”, and if your job is to trade bonds, go for it! But if your job is to grow long term compounded capital, we strongly argue that the ‘fixed income ship’ has sailed.

Should the Federal Reserve, and their committed pack of followers, successfully carry on their efforts, yields and carry will not keep pace with inflating assets. Should the Fed, either voluntarily or involuntarily, cause/allow the bubble to burst, these positions will not protect growth assets. We don’t know if carry will make or lose money in any given period, but we do know that concavity is the bane of compounding. Investors have to consider the opportunity costs in holding instruments that are losing out to inflation while providing no relevant risk mitigating benefit. It is dead capital.

As we keep harping on about, this has obvious implications for, give or take, the entirety of global savings that is managed in some sort of balanced portfolio context. We saw the below great visualization by the brilliant Gerard Minack in a recent John Authers’ Bloomberg Opinion article. As the dots have progressed up and to the left over time, 60/40 balanced portfolios have enjoyed the truly perfect environment. But what now? In the Minack piece, he discusses the linkage between historically high equity valuations and low interest rates. We might argue that the bigger takeaway from his lovely picture is that there is no upside in holding fixed income securities and, even more obvious, one is not a risk mitigant for the other.

Figure 7: The Golden Era of the Balanced Portfolio

Source: Datastream, Shiller, GFD, Standard&Poor’s, BLS; Minack Advisors

Many of you will have seen us draw the below pictures on any available whiteboard when we’ve met face-to-face (the good old days!). We use this as a visual example of the convexity difference between the linear relationship of a volatility swap to changes in vega versus the convex relationship of a variance swap (volatility squared). In the lingo of mathematics, it is referred to as Jensen’s Inequality which states that “the convex transformation of a mean is less than or equal to the mean applied after convex transformation”, and naturally the opposite for a concave transformation. We like to say that these two pictures clearly lay out the basic problem with BIS Regulatory Guidelines (Basel 3 now) and as such the global banking system, but most regulators we show it to don’t seem to get it. The gist being, that if your Normal Distribution’s two or three standard deviations is between -5 to +5 (the first chart), your “probability” based risk methodology tells you that being long the linear read function is higher expected return than being long the convex blue function, or alternatively that being short the blue line is less risky than being short the red line, or worst of all that being long the red and short the blue is “riskless” carry! Naturally, it will always turn out eventually that the interpolation of some short data series into a Normal (Gaussian) Distribution is wrong, and the tails are much fatter than implied by the two standard deviations.

Figure 8: Jensen’s Inequality (Vol vs Var)

Source: Convex Strategies

We gave a real-world example of this sort of payoff way back in our March 2019 Update, comparing a Convexity Swap to a St. Petersburg Paradox game: https://convex-strategies.com/2019/05/06/risk-update-q1-2019/.

We can use our scattergram tool to construct similar looking portfolios. Constructing the read lines as equities and fixed income, for 2020 when the range expanded to the -15 to +15 metric, a portfolio of 10% SPX and 90% Fixed Income (Bloomberg Barclays US Treasury Total Return Index) (it takes that much fixed income to offset the equity losses in the down market) creates a parabola that looks very much like our red line. You could think of this as being equally weighted to ongoing asset bubble versus a bust of said bubble. Then if we follow our standard reallocation rule of “take your Fixed Income allocation and split it 50/50 between Equity and Long Vol (CBOE Eurekahedge Long Volatility Index), then 2x lever the Long Vol”, you get 55% (10% previous allocation +45%) and 90% Long Vol (45% 2x levered), we get a parabola that looks very much like our blue line. It is like magic!

Figure 9: Scattergram Jensen’s Equilibrium: Balanced 10%/90% vs BarBell 55%/90% Jan-Dec 2020

Source: Bloomberg, Convex Strategies

We can, of course, look at it in compounded space as well.

Figure 10: Balanced 10%/90% vs BarBell 55%/90% Jan-Dec 2020

Source: Bloomberg, Convex Strategies

From the theoretical to reality. One need not guess, speculate, or predict whether we get boom or bust – just build a convex portfolio. The obvious pushback now is “sure, but what about the years prior to 2020, when we would have been much more in the -5 to +5 range, wouldn’t we have been better on the red lines?” We showed back in our August 2019 Update (https://convex-strategies.com/2019/09/30/risk-update-august-2019/)how the period of greatest relative performance of our Fixed Income proxy over our Long Vol proxy commenced in May 2012. So let’s recreate our parabolas and compounding tracks going back to that date.

Figure 11: Scattergram Jensen’s Equilibrium: Balanced 10%/90% vs BarBell 55%/90% May 12-Dec 20

Source: Bloomberg, Convex Strategies

Figure 12: Balanced 10%/90% vs BarBell 55%/90% May 12-Dec 20

Source: Bloomberg, Convex Strategies

The scattergram again is a near perfect representation of our Jensen’s Equilibrium theoretical example. The bulk of the occurrences take place in the narrower range, with virtually all the tail appearances taking place in 2020. Despite the narrower range from May 2012 up through 2019, there is still sufficient range of outcomes that our convex BarBell portfolio tracks at slightly better compounded returns than the risk equalized Balanced Portfolio, right up till 2020 where the more convex portfolio does what it is built to do.

We need to remember that we have constructed the BarBell portfolio to give equal weighting to boom and bust positions but to get equivalent sensitivity in the balanced portfolio we would have needed to own $9 of bonds for every $1 of equities. That weighting towards bonds gets bigger and bigger as interest rates get lower, and then becomes just flat out a bad idea (see Mr. Minack’s chart above). We say it over and over again. Own the things that are benefitting and participating in the inflation-fueled upside and stop investing in bounded upside, correlated downside, carry structures. It is hard to imagine, but it appears as though the legendary structured product buying Korean retail investors have been listening to us! What was the best performing equity index from the lows in March? Not S&P, not Nikkei and not even Nasdaq. You guessed it, Kospi! Figure 14, the longer-term chart, is an eye-opener.

Figure 13: Kospi2, NKY, SPX, NDX Indices 23 Mar 2020 – 31 Dec 2020

Source: Bloomberg

Figure 14: Kospi2 Index 2011-2020

Source: Bloomberg

Not quite Bitcoin, but pretty good asset inflation. Why do we say the Korean retail investor seems to have been listening to us? Because the guys that are notorious as the biggest vol suppliers in the business appear to have finally decided to stop chasing negatively asymmetric short vol carry (autocallable flows are way down) and to just outright get long! Meanwhile, local institution and foreign investors have decided to get out. Quite a stunning chart on the breakdown in flows. One wonders if it has anything to do with skin in the game? There is similar talk along these lines with the explosion of retail Robinhood traders in the US market and their appetite for call option buying.

Figure 15: Breakdown of Inflows to KOSPI

Source: JP Morgan, Bloomberg

We don’t know if central banks will fail or succeed in their ambitions to inflate asset prices. We do know that there is a whole bunch of stuff that we would classify as historical extremes; debt, deficits, interest rate levels, asset purchases by central banks, equity valuations, tech company market cap, leverage, unemployment. Extremes tend to beget extremes, especially in the world of money creation. The expansion of credit is inflationary and the existence of credit builds deflationary risks. Eventually more Snickers Bars just isn’t a good solution for the fatigued and bloated patient. In the meantime, build portfolios that participate in the upside, and have efficient protection on the downside, portfolios that are built to compound.

One last visual. If you can press yourself to imagine that there might still be some information value in interest rates (that takes a pretty serious suspension of disbelief in today’s world), the below chart of the slope of the US yield curve could be saying something about the market’s thoughts on inflation and cycles. The two dotted vertical lines are from August 2007 and March 2020. Maybe the powers that be have indeed ended it, but thus far throughout history it has always been a cycle.

Figure 16: US yield curve slope. 30y-2y and 10y-2y

Source: Bloomberg

Read our Disclaimer by clicking here