Allison's Ode to the Second Moment


Welcome to the 47th issue of Allison’s Ode the Second Moment, a newsletter whose name honors volatility but still understands there are other kinds of risk, too.

What Is Risk?

Last time, I wrote about what we mean by risk-free—the opposite of risk. This week, let’s talk about what risk is and how we measure it. Robin Wigglesworth has a nice essay in The Financial Times about the “volatility virus” that “infected Wall Street.”

I am not generally a fan of stories about a common financial tool that some people use improperly, because they either weren’t trained or were just greedy, as an example of yet another hair-brained-yet-Nobel-Prize-winning idea that destroyed the entire economy, or in this case, just the “wide carnage” we all experienced a few weeks ago. But this essay does raise some important questions about how we think about risk.

Wigglesworth explained how volatility became the go-to risk measure in finance. It is, if you assume a vaguely normal distribution, the range of things that might happen to an asset price most of the time. It does not tell you much about tail risk, which is an improbable and extreme event. But that does not mean volatility it a terrible way to measure risk.

When you go to the airport, odds are your estimated travel time is based on a range of typical traffic conditions. You don’t account for a tail-risk event, like a 300-car pile-up, every trip. If you did, you’d spend way too much of your life in airports. In finance, as in life, we often make a risk decision based on the range of things that are most probable.

A simple, quick, and easy to replicate measure of risk is equally valuable in financial markets, even if it is incomplete. Risk is both the probable and the improbable. Good practitioners recognize the limitations of volatility as a measure of risk, insure against downside tail risk, and limit their leverage.

Personally, I think price fluctuations in markets are healthy. Stocks are risky assets. What was weird was the lack of price volatility the past several years. But markets are weird. The ability to estimate volatility reduces risk, but it never eliminates it. I suppose maybe it is sad some naïve investors lost money when they bet markets weren’t risky anymore. But don’t blame Harry Markowitz for their mistake.

Markets are uncertain. Finance has tools that attempt to measure the unmeasurable, and most of the time these tools work. But using a tool well requires an understanding of its limitations. When something goes wrong, it is worth asking, Did the tool fail or did the investor use the right tool for the right problem? You don’t use a screwdriver to hammer a nail, you don’t diversify to hedge systematic risk, and you don’t use volatility to measure tail risk.

A Cautionary Tale of Not Measuring Risk at All

An imperfect measure of risk is better than ignoring risk all together. Mary Walsh has a fantastic story about Oregon pensions. The story was very popular, which made me very happy because it shows that the world is full of Pension Geeks...or just concerned taxpayers.

Mary describes a DB plan where the benefits were based on earnings from any job, not just the state job. And it was not just the DB plan that overpromised:

For decades, every public worker got a simulated tracking account. It was credited with 6 percent of each paycheck, then left to compound at a predetermined rate.

In the early years, a low rate was used because the pension system invested in bonds that didn’t yield much.

But in the 1970s, lawmakers started nudging the rate up, eventually to 8 percent. Then, the system’s trustees decided 8 percent should be a guaranteed minimum. In years when markets produced higher returns, the accounts compounded at those rates, after money-management fees. During the 1990s bull market, accounts compounded by up to 21 percent a year.

When workers retired, their employers were required to “match” the account balances, doubling them. Then PERS would base the pensions on the total.

Today, services, especially for children, are being cut to pay for pension benefits.

Other states have made equally generous guarantees. It is easy to single out state politicians who had every incentive to promise more later and underfund today. But humans have short memories, and after a few years of low volatility in the markets, people believed that stocks were risk-free. When states made these crazy promises in the 1980s and 1990s, they never imagined interest rates would fall and stay low. They never imagined people would move out of their state or there would be less growth. They did not under-estimate tail risk, they assumed there was no risk at all.

Meanwhile, The Economist, is not worried about the government's liabilities. But instead of counting on high rates, as pension plans did, they expect interest rates to stay low. They argue even if rates do go up and growth does slow, the Fed can monetize and count on growth to make the problem go away.

America faces a big challenge balancing its books in the long term. If it does not, interest rates must soar eventually. But countries usually have fiscal wriggle room as long as they grow, in nominal terms, at a rate higher than the interest on their debt. America remains well within this comfort zone. Rates may rise a little more, but that would give the Fed welcome room to loosen policy the next time recession strikes. Pessimists who thought rates would never rise were wrong. Today’s bond-market doomsayers probably are, too.

So much for tail, or even non-tail, risk.

Here's one big risk that might work out

As regular readers know, I like to find risk in unusual places. I recently looked at the cruise industry. Taking a cruise is a hedge. You give up the upside of a thrilling adventure on the trails of the Andes or camping on the beach in Bali, but you get a fairly predictable vacation that will be sort of fun for everyone. But there is still tail risk, like being on a ship that gets stuck in the middle of the ocean with no sanitation.

I interviewed Carnival CEO Arnold Donald. His hedging strategy took him from poverty in New Orleans to the C-suite. It also turns out Carnival is doing something that offers insight into the future of work. Often, we learn the most about technology outside Silicon Valley.

Carnival’s Princess Cruise line is currently testing the Medallion. Donald poached John Padgett from Disney to create it. Padgett is the man behind the Magic Band, an early example of wearable technology that streamlined the Disney vacation. Now at Carnival, he developed the Medallion. It takes the Magic Band a step further. It tracks you throughout your cruise and uses AI to anticipate what you want before you know you want it. Say you had a martini on the sun deck last night, the Medallion uses that information to suggest you try snorkeling.

We assume AI will replace humans, but that is still conjecture. Technology in the past augmented human labor instead of replacing it. The Medallion is an example of labor-augmenting AI. Carnival still needs human crew members, but they are AI-powered. Carnival’s goal is to offer a luxury level of service at scale. AI technology does this by turbocharging human connections instead of replacing them. Crew members know your name and what you like, even if they’ve never met you before. It sounds a little Orwellian, but it also offers a glimpse into what AI could mean for the future of work.

Maybe massive job loss is one tail risk we needn't worry about.

In other news

The Puerto Rican control board has a plan that involves cutting pensions and other spending. Larry Summers probably thinks they are being too optimistic and that the plan does not offer enough financial support for Puerto Ricans.

The IMF is worried about debt. And lots of people are worried about the yield curve inverting, a consistent predictor of a recession. But the IMF, now featuring Tobias Adrian (someone who really understands long-term interest rate risk) speculates yield curves might be flatter because the ever-present promise of more QE shrunk term premiums.

Until next time, Pension Geeks!