Rather than the gossipy crap I’ve heard all month, here’s some deserved props, appreciation, and a little deconstruction for Bill Gross.
Bill Gross deserves props, appreciation, and a little deconstruction for five solid decades.
Bill Gross has forgotten more about bond trading than most bond traders will ever know.
A sample of his artwork: implementation of portable alpha and smart beta.
In the post-crisis world, stability—not boldness—became the cardinal virtue. Bill’s basic skill set is less appreciated in today’s markets.
His response was to open an unconstrained bond fund.
His poor couple of years at Janus Henderson is due to his investment approach. The same approach also reflects his genius.
His monster trade at Janus was a variation of the widowmaker. It killed him.
Bill Gross announced his retirement from Janus this month due to a bad couple of years running an unconstrained bond fund, both on a total return basis and an AUM basis. After nearly 50 years of solid and often brilliant performance, it is to be expected. Markets reinvent themselves in a wholesale fashion every three to five years, so the fact that he was in the industry pole position for so long is remarkable. It is ironic that his poor couple of years at Janus Henderson is due to his investment approach, because it also reflects his genius.
Bill Gross has forgotten more about bond trading than most bond traders will ever know. He had a magnificent insight, which seems so commonly obvious that we are hardly able to give it the weight due to it: total return bond portfolio management. We certainly do not appreciate the enormous effort and thinking its attainment cost.
Before Bill Gross, bond investing was governed by three imprecise principles they called: safety, marketability, and satisfactory yield. We call these today credit risk, liquidity, and collection of carry. Carry is the differential between forward and spot yield of the bond, and it indicates the profit and loss an investor realizes by holding the bond against cash for a defined time. These principles were effective, but imprecise guides to money management.
Bill Gross didn’t translate these principles with mathematical precision. He hired the people who did. He brought in the brightest minds in mathematical finance—guys that homesteaded stochastic processes in an asset class that remains dominated by phone calls between traders and their lawyers. After the deal was settled, Bill Gross figured out that risk could be decomposed, bought, sold, and hedged. . He then turned mathematical constructions into a practical and usable framework. Underlying this is an essential interconnection tying all asset classes together that is worth demonstration.
An asset price is described completely by a deterministic equation. This is called a risk-free asset. If a market has a riskless asset, then its price B follows the ordinary differential equation
where r is a continuous nonrandom short rate of interest. You can think of this as a share in a money market fund.
A riskier bond (or other asset) has more complex dynamics simply by virtue of having credit, market, or interest rate risk. These risks introduce a random component into the price dynamics.
In financial models, this randomness is formulated in the differential equation
Where r is the short rate of interest associated with the risk-free asset, and st is another continuous nonrandom function of time. This equation is called arithmetic Brownian motion, commonly rewritten as
Where the term Wt is the random component, conventionally the standard Weiner process.
Stocks, bonds, and money markets are all connected in the following way.
Thus the drift component of a random asset price is equal to the short rate r. This implies that defaultable bonds and stock prices and their total cumulative returns are linked to risk-free money markets. In fact, provided there is a risk-neutral measure, bond returns are anchored by the return on risk-free bonds. Excess returns are due to collected risk premia, and the associated risks must be managed by buying or selling exposures.
Bond management has always weighed carry against risk. Pre-Bill Gross, evaluation was carried out with balance sheet review, gut instinct, and rating agency assessment. Still is to some extent. But that is understood as only a fraction of the task. The accepted way to assess risk now is to also look at tail events and their impact on a portfolio. Value-at-Risk on risk-weighted assets is that standard yard stick, indicating the worst loss after a period for a given confidence interval. Try as we might, we look at bond management performed before Bill Gross sympathetically, but as the work of half-savages.
His perspective is summed up in this quotable gem: “I always said a great portfolio manager would be one-third mathematician, one-third economist and one-third horse trader—it's the horse trading that gets into human nature.”
Each of these skill sets contribute to his total return concept in fixed income. The trick was and still is how to best apply them in the right proportion to the asset classes in credit land. A decent part of this entails rigorously applying basic and sophisticated arbitrage pricing bond math. Just as important is an understanding of how news flow, policy, and business cycle mechanics impact and are impacted by financial markets. Lastly, you should be a good horse-trader: keep the nerve to assess risk, press even small advantages, and aggressively power through illiquid, legally complex transactions. Gross’ total return genius was developing portable alpha and refining smart beta.
The basic idea behind portable alpha is finding source of return uncorrelated to a benchmark index. A simple example of a decently complex process involving a derivative overlay: long a benchmark like S&P 500 for return, juice returns with a higher beta component like small caps, then hedge the small cap exposure to manage the overall beta down to the benchmark target. Bill Gross made terrific hires like Myron Scholes, Vineer Bhansali, and Graham Rennison to apply and customize derivative overlays on all kinds of illiquid bond portfolios.
Smart beta is a different animal. It decomposes portfolio return into multiple factors that can be bought or sold as another way of buying or selling risk exposures. These factors can be used to reconstruct whole beta or weighted to add a tactical tilt to a beta position. While there are some variations in how they are formulated, the basic factors are:
Growth (solid performer since 2009, with 2015 and 2018 as exemptions)
Momentum (solid performer and even decent in 2018)
Low volatility (decent in the 2018)
Value (rotten performance for years)
There are two families of factors: those that depend on mean-reversion like value and low volatility, and factors that are not mean-reverting, like growth and momentum. With them, one can express precise a macro-view on risk and have plenty of liquidity simultaneously.
Smart beta and portable alpha were bold, innovative, and different. In the post-crisis world, stability—not boldness—became the cardinal virtue. I suspect Bill’s core skill set was less valuable.
His response was to open an unconstrained bond fund.
His last monster trade was on BTP-Bund convergence. It made sense, and still does. If the economy recovers BTPs would catch bid and Bund yields will rise on lowered safe haven demand. Convergence. If the economy doesn’t recover, then fast money would front-run ECB purchases of BTPs, driving yields down. Convergence.
It made sense, but it didn’t work. Instead, the positive effect of an expected economic recovery was dwarfed by the fear that the ECB would be ending sovereign debt purchases, if only because they were running out of assets to purchase. Combined with this, the market evaluated Germany as a true sovereign in the “central bank backstop” sense. Italy was not seen as such a beneficiary. The result was a bloodbath. Gross saw an estimated one day MTM loss of $63 million on AUM of $2 billion, that spike is likely enhanced by the unwind of his own trade.
The BTP-Bund spread is a very useful risk-on, risk-off indicator from mid-2018. No doubt that Bill Gross will be right about BTP-Bund convergence. It is a high-convexity trade long on the EU project, and there are so many sunk costs embedded in the EU project that it will not die easily.
But the trade had a short bunds leg, and you seldom win with these kinds of widowmakers.