Lost decades: Risk scenarios also matter on the way back up

There's something a lot worse than being down 35%, and that's being down that much and then going nowhere for a decade or more.

If a client has 15 or 20 years to reach their investment goal, no worries about the market’s ups and downs, right? Here’s some background to think about.

Bad things happen from time to time. Over the last 50 years there have been five instances when the market has dropped more than 35% — pretty much one time each decade. That can put a real damper on a client’s goals if it happens without leaving time for their portfolio to recover.

Monte Carlo simulations that are used to project where portfolios might end up don’t reflect these sorts of moves. They are based on a normal distribution, where the odds of having these drops are less than one in a hundred — that is, they might come around once every 5,000 years. A normal distribution leads to a smooth incremental path without large jumps

That’s bad enough, but there’s something a lot worse than being down 35%, and that’s being down and then going nowhere for a decade or more. The phrase “Lost Decade” entered the popular lexicon when the Japanese market floundered after a huge drop in the early 1990s. (Actually, it didn’t stop after 10 years; better to use “Lost Decades.”) The U.S. equity market also has had lost decades — and then some — where the market sold off and then took well over 10 years before recovering to its previous level.

A TALE OF THREE DRAWDOWNS

Three of these periods occurred from 2000 to 2013; 1968 to 1983; and 1937 to 1950. Adding in 1929 almost wouldn’t be fair. These weren’t one-shot drops. Each reflected the accumulation of a sequence of bad events. For example, in the 2000s, there first was the rapid deflation of the dot-com bubble, then earnings adjustments from levels of “irrational exuberance,” then the shock of accounting frauds at Enron, WorldCom and others. Follow that with the trifecta of the subprime mortgage meltdown, Bear Stearns, and Lehman Brothers that propelled the 2008 banking crisis, and the subsequent “PIGS” credit crisis (Portugal, Ireland, Greece and Spain).

A curious but not reliable factoid is that in each case it was just under 20 years from the end of one of these to the start of the next. That is, the first ended in 1950 and the next started in 1968; then that one ended in 1983 and the next started in 2000. So, no worries, the next round won’t be until 2030 or so! (Just kidding.)

WHEN BLACK IS THE NEW RED

The critical point for clients to keep in mind is that things are worse than they may seem if their portfolio’s flat for 13 or 14 years. Their investment goals are based on an assumption that the portfolio will grow. If a client is sitting at the start of one of these lost decades with a portfolio of $1 million in equities and is looking 15 years out to tap that portfolio for some major spending, a 7% per annum capital market assumption will project a terminal value for the portfolio of $2.75 million. If the stock market is flat for 13 of those years, the client will end up with something more like $1.15 million. While the client is nominally in the black, they’re likely to feel as if they’ve experienced a ’29-style crash because they’re down 60% from expectations — keep them away from high windows.

How clients fare can seem like a matter of luck of the draw in terms of timing. And bad timing doesn’t have to involve buying at the top. It took more than a decade for the market to head back up for good after February 1961, as well as after September 1996. Both 1961 and 1996 were unremarkable years. Indeed, in both cases the market was on a upward trajectory; it wasn’t on the cusp of irrational exuberance or paralyzing recession.

No one wants their clients’ goals to be held hostage to luck-of-the-draw timing. So portfolios have to be constructed and their risk managed with these possibilities in the forefront. But before doing that, advisers must recognize that these are an unavoidable reality of the markets.

The bottom line is that long-term goals don’t confer immunity to market risk. On a time horizon of 50 years or more, maybe. But if it’s 10 or 20 years? No. Clients’ risk is being underestimated if simulations don’t reflect the major market events and the flat-markets epochs.

SHORTCOMINGS OF MOST SCENARIOS

This is a major issue for scenarios. If they’re not looking at the path of the scenario, they’re not treating its risk. A scenario is not simply: If such and such an event occurs, then boom, equities will be down 25%. First of all, there’s no magic number. Maybe it will be down 25%, maybe 30%. Second, there’s a time dimension: Maybe things drop quickly, maybe gradually. For example, in the current milieu, it’s likely that if we have a retrenchment in technology, equities will drop quickly, whereas if inflation is the catalytic event, asset prices could move down more slowly.

Most important within this time dimension is the recovery period — the time it takes before the event recedes into history and the market is on its way up again. That can take years, with some false starts on the way.

This happens for two reasons. First, the market is weakened, and events that might not have normally mattered can weigh on returns. Second, as recounted above for the 2000s, subsequent events can come along that create a one-two punch (or sometimes, a one-two-three-four punch). This reality isn’t part of standard risk management practice today. Scenarios must be built with the time dimension included, because the issue isn’t simply how far down the market might go, but how long it takes to recover.

[More: How to make risk management work for advisers]

Rick Bookstaber is co-founder and head of risk at Fabric. He previously held chief risk officer roles at Morgan Stanley, Salomon Brothers, Bridgewater Associates and the University of California Regents, and served at the Treasury in the aftermath of the 2008 crisis. He’s the author of “The End of Theory” (Princeton, 2017) and “A Demon of Our Own Design.”