In his most recent annual report, Berkshire Hathaway chairman Warren Buffett described an event that occurred at a Berkshire subsidiary last year: Late one night, a fire spilled over from a neighboring business, resulting in significant damage to the Berkshire facility, forcing it to shut down. Fortunately, no one was injured. And as Buffett notes, the losses will be covered by insurance. But unfortunately, one of the company’s largest insurers was, as Buffett put it, “a company owned by . . . uh, Berkshire.”

The lesson: Even the most well-managed enterprise—and a company in the insurance business itself—can find itself the victim of unintended consequences.

In recent weeks, I’ve been thinking a lot about this story. While we all understand the importance of diversification to manage risk, sometimes things go wrong in unexpected ways.

In my view, the most unexpected aspect of what’s happened in financial markets this year is what happened to bonds. Traditionally, stocks and bonds have exhibited negative correlation, meaning that when one goes down, the other goes up, and vice versa. That’s what makes them such a powerful combination in a portfolio.

But this year, when stocks went down, most bonds started to go down along with them. It was like finding a hole in a life raft. While this has since started to reverse, I think it’s worth pausing to understand what happened, why it happened and what lessons we can learn from it.

What happened?

To start, let’s go back to February 20th, which was when the stock market started to falter. At first, bonds were behaving in character, rising as the stock market was falling. For the first few weeks, between February 20 and March 6, while the stock market was falling, the bond market rose—just as expected. But then, over the following week, as the stock market’s losses deepened, the bond market started to decline too. In all, the bond market lost nearly 9% in just that one week.

Certain areas of the bond market fared especially poorly. Over the one-month period between February 19 and March 19, short-term corporate bonds lost 12%, and high-yield (aka, junk) bonds lost 19%. Even municipal bonds lost 16%. This was probably the biggest surprise. During this period, in fact, the only winner was U.S. Treasury bonds, which rose 2%.

Why did this happen?

A number of factors came together at once. The first two are typical of any stock market downturn, while the second two are more unique to this year’s situation:

• Liquidity needs. At any given time, even in normal times, there is some number of people who need to withdraw cash from their portfolios. These include retirees and university endowments, among others. But when the stock market started to drop, these investors did the rational thing and sold their bonds instead of stocks, putting downward pressure on bond prices.
• Margin calls. One of the ways brokerage firms make money is to let investors borrow against their portfolios. When the market is going up, this may seem like a good idea. But when the market turns south, some of these investors receive a dreaded “margin call.” This is a demand from the broker to pay down their loan. Faced with this problem when the stock market is down, investors normally opt to sell bonds instead of stocks. This too contributed to selling pressure on bonds.
• Fear—some justified, some not. When broad-based quarantines went into effect this month, many bondholders began to worry about the impact on companies and municipalities. This also contributed to selling pressure on bonds.

• Tax law changes. At the end of 2017, the Tax Cuts and Jobs Act had the effect of raising taxes for many people, especially high-income individuals in states with high state and local taxes. The result was to drive many of these people into the municipal bond market, where they could earn income free of federal taxes. As a result, many municipal bond funds swelled in size over the past few years. That was fine until the aforementioned fears arose, and that caused an unprecedented stampede back out of bond mutual funds, forcing bond fund managers to unload their holdings at any price. The selling pressure was literally off the charts, as shown in Figure 2 of this report.

While it is difficult to quantify the relative contribution of each factor, it was the combination, all at once, that caused the bond market to seize up in ways that, according to one analyst, hadn’t been seen in forty years. Fortunately, the Federal Reserve, using its ability to effectively print money, has since stepped in, and that has led to a significant, though not complete, recovery in most areas of the bond market.

What can we learn from this?

I see three useful lessons from episode:

There’s nothing wrong with a belt and suspenders. I often talk about the danger of Recency Bias, which is the natural tendency to extrapolate from recent events. It is dangerous because, until this year, the stock market had gone up nearly every year for eleven years. The result was that many people got lulled into a false sense of security. This year’s market turmoil, though, is a good reminder that it’s okay to build a little paranoia into your portfolio. That might take the form of cash or a money market fund, or maybe short-term U.S. Treasury bills. There is no science to this, except to be cognizant that risks exist in categories that we may not even be aware of.

Remain diversified even when it feels unprofitable. Another truism about investing is that every crisis teaches us something new—usually the hard way. In this case, it was that unique combination of four factors—two old and two new—that caused the bond market to come unglued. Next time, it will be something else. All of this is a reminder that our only and best protection is broad diversification. One investor asked this week whether he should move his entire portfolio into U.S. Treasurys, since those are what have held up best this year. It was a good instinct. After all, in finance textbooks they characterize U.S. government bonds as a “riskless” asset. But my response was that nothing is guaranteed—not even Treasury bonds. In fact, you may recall the government shutdown in 2011 that caused U.S. government debt to be downgraded. Do I still feel very comfortable with Treasury bonds? Yes, absolutely. And it has certainly been the star this time around, but that may not always be the case.

Measure twice, cut once. If there is one company that suddenly everyone knows, it is Zoom, the videoconferencing company. It’s a great product, and the stock (ticker symbol ZM) has enjoyed strong gains this year—up 123%. But there’s another Zoom that has done even better. It’s an obscure Chinese company with no revenue that happens to be listed on the U.S. market, and with a much better ticker symbol: ZOOM. As a result of their confusingly-similar names, this other Zoom’s stock, which in the past typically traded for about a penny a share, has shot up nearly 900% this year. The lesson: If you’re making changes to your portfolio in this environment, go slow and be careful of making decisions under stress.