Year-to-date, stocks are down almost 20%. And bonds, surprisingly, have lost money too—down about 10%. At times like this—when the headlines are almost all negative—the standard advice is to avoid panicking, to stay focused on the long term. I agree with that, and indeed the data are clear: Investors who attempt to time the market with “tactical” trades often suffer whipsaw. But that doesn’t mean we should bury our heads in the sand either. There are at least four situations in which it might make sense for investors to sell when the market is down.

Cash flow. The first—and most obvious—reason you might sell stocks even when they’re down is to meet cash needs. In general, I recommend that investors hold at least three years of expenses in cash or bonds, if not more. But what if your portfolio wasn’t structured that way before the market dropped? In other words, what if you find yourself now with cash and bond holdings that are too light to meet upcoming expenses? It might seem like an unpleasant prospect to begin selling now. That’s understandable. Ideally, we would only sell when the market is strong. Nonetheless, I wouldn’t hesitate to sell now—for a few reasons.

Yes, the market is down. But relative to past downturns, things really aren’t so bad right now. The S&P 500 is down 18%. While that might sound like a lot, that is relative to a near all-time high—after more than a decade of almost continuous gains. Compared to just three years ago, for example, the S&P 500 is up 37%, and bonds are down only 3%. So a sale at today’s prices is hardly a fire sale. And if selling some shares at today’s prices would help you to build up a cash reserve, I wouldn’t hesitate. And then you could sleep easy, even if the market dropped further.

Sub-optimal holdings. In my view, there are four types of investments that are less than ideal. First are highflying investments—things like speculative technology stocks. Second are stocks or funds that simply have a lackluster outlook. Unlike highflying stocks, lackluster ones probably won’t drop too much. Instead, the risk is that they simply might not gain very much. Many banks and old-line industrial companies fit in this category. Third are high-cost mutual funds. The data there is clear: On average, funds with higher expenses have underperformed funds with lower expenses. The fourth category includes private funds, where transparency and liquidity are generally low. As I’ve noted before, publicly-traded stocks rarely go to zero. But I’ve seen private fund investments lose all their value more than once.

If you hold investments that fit any of these categories, my advice would be to look for a way to get out. When the market is high, such an exit might carry a tax cost. But with the market lower today, you might be able to exit a sub-optimal investment with little or no taxable gain. The proceeds could then be reinvested immediately into investments with a more promising, or less risky, outlook. The nice thing about a swap like this is that you wouldn’t be a net seller while prices are low and wouldn’t be changing your asset allocation. You’d simply be upgrading your portfolio.

Tax losses. Even if your portfolio isn’t saddled with any sub-optimal holdings, market downturns offer another potential benefit. Tax-loss harvesting is the strategy by which an investor sells an investment—usually a fund or ETF—then immediately purchases a replacement with the proceeds. That provides a valuable and flexible tax benefit. First, the investor could use that loss to offset capital gains on other sales. Or if there are no gains, up to $3,000 of the loss could be applied against ordinary income, such as wages. And to the extent that there are unused losses beyond that, those could be carried forward to future years.

Here’s what makes tax-loss harvesting so powerful: When you do a swap like this, you can collect the tax advantage without substantially changing your portfolio. Under IRS rules, you can’t sell an investment just to book the loss, then immediately buy it back. In fact, the rules say that you can’t buy back anything that is “substantially identical.” Otherwise, it’s called a wash sale and the tax benefit is voided.

You can, however, buy back an investment that is similar—just not identical. Suppose, for example, you own a fund that tracks the S&P 500. If you sold that, you’d have several options for replacing it. You could buy a total-stock-market fund. That sounds like a different investment, but it has an overlap of more than 80% with the S&P 500, so its performance shouldn’t be all too different. And while the IRS has never precisely defined its “substantially identical” litmus test, most advisors agree that a swap like this would not run afoul of the rules.

Alternatively, you could swap into a fund that follows a large-cap index other than the S&P 500. Vanguard, for example, offers a fund that tracks the CRSP large-cap index (ticker: VV) and one that tracks the Russell 1000 large-cap index (ticker: VONE). If you were to compare the historical performance of these two funds to an S&P 500 index fund, you’d find them to be very similar—but not identical.

Another key point about tax-loss harvesting: By immediately replacing the investment you sell with a fund that is similar, you again are not a net seller. So you shouldn’t feel you are doing something unwise by selling while the market is down.

Imbalances. Another reason you might consider selling even when the market is down is to correct a significant portfolio imbalance. These imbalances can take many forms. Some investors have a concentration in a single stock. I often see accounts, for example, that have disproportionate holdings in Apple or Amazon. In both cases, these imbalances are the result of great performance in recent years. But nonetheless, looking forward, they represent risks. Imbalances can also find their way into portfolios via mutual funds that are concentrated in certain parts of the market—technology stocks, for example. Imbalances can also crop up on the bond side.

In each of these cases, there might be nothing wrong with any one individual holding. The problem, though, is the risk they present in aggregate. Harry Markowitz, the father of Modern Portfolio Theory, said it best. In his initial work back in the 1950s, he used railroad company stocks to explain the concept of diversification. There’s nothing inherently wrong with railroad stocks, he explained. But if a portfolio consists of only railroad stocks, that’s a problem. It’s a problem because companies in the same industry are often impacted by the same economic factors. As a result, a portfolio may only appear diversified. But if too many of its holdings are concentrated in one area, it may be more risky than it appears.

“One hundred securities whose returns rise and fall in near unison afford little more protection than the uncertain returns of a single security,” Markowitz explained. If you identify an imbalance like this among your holdings, that’s another situation in which it may be worth selling even when the market is down, and even if it will generate some tax, to correct it.