The poor risk management at Silicon Valley Bank is stunning. But what went wrong is actually pretty common. There is often confusion about what makes a safe asset versus a risky one, and guessing wrong is often at the core of financial blow-ups. Safe assets, as we properly define them, are the backbone of financial markets. They are how we price and measure risk.

But knowing what makes an asset safe is not always so simple, and it will differ based on circumstances. Odds are you have the wrong safe asset in your retirement portfolio right now.

We tend to think of government bonds as safe. They are liquid, meaning you can sell them pretty easily and quickly. Debt-ceiling theatrics aside, it’s a safe bet the U.S. government won’t default on them. This is why bank regulation considers these assets to be low-risk. But depending on your financial situation, not all bonds — even U.S. government ones — will be safe.

Silicon Valley Bank offers us all a cautionary tale. It had short-term liabilities, or deposits, which it financed with long-term debt — Treasuries and mortgage-backed bonds. The issue wasn’t liquidity since it’s easy to sell a U.S. Treasury to raise cash. The problem was that the value of the Treasuries fell when rates went up. Longer-term bond prices move a lot when rates rise or fall, while the value of the bank’s liabilities (the deposits) didn’t change — and suddenly all their depositors wanted their money. All banks have a similar mismatch with their assets and liabilities, but the nature of Silicon Valley Bank’s deposits, which were large, mostly uninsured and mostly from tech firms that would need their money once interest rates increased, made it especially risky.

In hindsight, this was obviously terrible strategy. Not only because the bank doubled down on its bet that rates would stay low, but that it made this bet when interest rates were at record lows, which were never going to last.

But to be fair, no one can predict the future, and it had been a long time since bond prices fell by a large amount. Bond prices had been trending up for the past 40 years. It had become a widely held belief that we had entered a new era of forever-low yields and if rates did go up, they wouldn’t rise by too much. Perhaps some people assumed that the introduction of Quantitative Easing as a policy meant the Federal Reserve was committed to keeping the entire yield curve low forever — this seemed like a reasonable assumption based on Fed policy during the past 15 years.

Pension funds in the U.K. also bet on low rates forever, and we’ll soon find out who else did, too, because a changing rate environment reveals where all the bodies are buried. If you assume rates will stay low forever, short- and long-term bonds are almost interchangeable from a risk perspective. They appear equally safe, with the only obvious difference being that longer-term bonds offer slightly more yield.

The reality is that predicting interest rates is hard and there are no guarantees. And whether long-term or short-term bonds are safe depends on the nature of your liabilities. The only safe strategy is to hedge interest-rate risk by, for example, buying bonds with a similar duration as your liability.

Hence Silicon Valley Bank needed to shorten the duration of its asset portfolio. For SVB, short-term debt like 3-month or 1-year Treasury bills (or interest rate swaps to get the same effect), whose value doesn’t change much when rates go up, would have been the safe asset. It would have sacrificed some yield in exchange for the lower risk of more stable prices.

If you are saving for retirement, you face the opposite problem. Short-term assets are your bigger risk. Your liability is the need to finance spending in your retirement. This is a long-term liability, as you can expect to live at least 15 years after you retire.

So think of the money you’ll spend in retirement — let’s say $50,000 per year protected from inflation. Financing this is comparable to issuing a long-term bond that pays out $50,000 (with an inflation adjustment) per year for 15 years. So if you want to ensure that you have $50,000 to spend in each of those 15 years, you can buy a bond portfolio that has the same duration: It will make $50,000 inflation-adjusted payments to you each year no matter what happens to interest rates.

Two years ago, a bond with this duration would have cost about $803,000. Last week it cost $665,000, almost 20% cheaper. That might seem good now, since the rising bond yields essentially made future spending cheaper. Here’s another way to look at it: Suppose you saved $803,000 and used that money on day one of your retirement to buy a bond that would finance your spending for the next 15 years. Two years ago that money would buy you $50,000 in annual inflation-protected income, but with recent bond yields, that same amount of money would buy $60,000 per year. Your income went up because interest rates have gone up.

But here’s the thing that so many people overlook: Rates also might go down in the future — or move up and down more frequently. Investing in short-term bonds amounts to not hedging your risk, and leaves people vulnerable to big swings in retirement income, sometimes good, other times bad.

The problem is that in the retirement industry, short-term bonds are considered the safe asset because their prices are stable. Many savers invest their retirement account in target-date funds, which move your portfolio into short-term bonds as you approach retirement to ensure your asset balance stays steady. The thinking goes that you can then take out a fixed percent each year. But this strategy means much lower returns (less income), more risk and no inflation protection. Most people can’t afford that choice, and it’s extremely risky in today’s inflation environment.

So if you want safety and predictability, you need to hedge risk and invest in longer-term inflation-protected bonds before and then during your retirement. Unfortunately, the odds are we still haven’t learned this lesson. In part because retirees came out ahead this time when rates went up, even though that might not be true in the future.

Such is the nature of risk. And it shows why it’s easy to misjudge what is risky and what is risk-free — and why hedging is so valuable.