It’s that time of year again, when Wall Street strategists begin publishing their market forecasts for next year. If you’re wondering whether to put any stock in these glossy publications, here’s my recommendation: Think back a year to the forecasts issued at the end of 2019. Did any of them predict that a virus would come out of left field, throwing the economy into recession, triggering a bear market, and killing more than a million people worldwide?

This is just the latest illustration of the reality that, for better or worse, personal finance is not a science. To be sure, finance involves mathematical analysis. But anyone who’s being honest about it will acknowledge that, in most cases, financial analysis requires making one or more assumptions. The result, as you might imagine, is that it’s easy for an analysis to provide two conflicting answers to a given question.

Consider a simple example: If you want to make a charitable contribution, should you do it now or wait until 2021? The answer is that it depends on where tax rates are headed. If you believe the new administration will raise rates, then you’d certainly want to wait (because your deduction will be worth more when rates are higher). But if you think they won’t raise taxes until the economy improves, then you’d want to go ahead and claim the deduction this year.

These kinds of questions can be frustrating. And in the absence of a crystal ball, many people rely on one of the following approaches—neither of which is ideal—for decision-making:

Option 1: Do nothing. When an analysis doesn’t provide a clear answer, it’s natural to want to shelve a decision. In a lot of ways, that makes sense: If you feel like there isn’t a tangible basis for making a decision, then maybe it’s best to defer it.

Option 2: Do something. If you don’t like the idea of shelving a decision, but you still don’t have enough information to go on, then you might just take a guess, or go with your gut, on the basis that any decision is better than indecision.

I’d like to offer another alternative: When there is no quantitative way to make a definitive decision, often the best approach is, for lack of a better term, to split the difference. While this might sound informal and not a rigorous way to make important decisions, I believe it’s better than the alternatives. Where can you apply this split-the-difference principle? Here are some examples:

Roth vs. traditional 401(k): If there’s a Roth option in your 401(k), is it worth foregoing the tax deduction to make Roth contributions? And if so, should you allocate your entire contribution to the Roth account, or just a portion? This, of course, depends on a comparison between your current income (which you know) and your future income (which you can only estimate) and whether Congress will raise rates in the future (which no one knows). My advice: Even if you’re in a high tax bracket today, you might still split the difference with your contributions.

Roth conversions: If you’re currently in the early years of retirement and have an IRA you’re considering converting to a Roth IRA, you’ll be asking similar questions about tax rates. But it might be more urgent. That’s because the tax cuts implemented by the 2017 Tax Cuts and Jobs Act aren’t set to expire for another five years, but Congress and the new administration might agree to raise them sooner. Will that happen? That depends on how quickly the economy recovers and on which party controls the Senate. And there are probably other political considerations. No one knows where this will all end up. In other words, tax rates might go up, making accelerated Roth conversions more attractive. But they might not, in which case accelerated conversions might end up costing you more. So again, a sensible strategy might be to convert some, but not all, of your IRA funds, and take it year by year.

Mortgage term: In general, if you go with a 15-year mortgage instead of a 30-year, you’ll pay a lower rate, and you’ll be done a whole lot sooner. But your payments will be much higher. How can you split the difference on this? One way is to take out a 30-year mortgage to preserve flexibility, but then commit to making larger payments in months when cash flow permits.

Paying off debt: Suppose you’re in a position to pay off your mortgage or other loans? With rates at all-time lows, is it worth it? This depends on what you would otherwise do with the funds. If you invest it in the stock market, your returns will probably be higher on average, but not necessarily every year. So it might be a good idea, or it might not. That’s why you might split the difference. Maybe use a third of your cash to pay down debt, invest a third in the market and keep the remaining third in the bank for flexibility.

Buying investments: This week, someone asked me to research options for investing in Israel. I found two funds that fit the bill. I liked them both, but they were different, and I couldn’t say that one was necessarily going to be better than the other. So I recommended buying both in a 50/50 split.

Selling investments: Do you own overpriced funds or underperforming stocks? Ideally, you would sell them, but that might entail a tax bill. Then again, the cost of underperformance might outweigh the tax bill. Fortunately, it doesn’t need to be all or nothing. You could sell a little bit each year.

Claiming Social Security: Do the math on Social Security strategies, and the answer is usually that high-income individuals should wait as long as possible to claim their benefit. But that assumes you’ll live as long as a life expectancy calculator predicts. My advice: In light of this uncertainty, it’s not heretical to at least consider claiming at 68 or 69.