Is buying random stocks and then selling off the ones that go down a good way to reduce capital gains tax?

You're intentionally investing in stocks that you hope will lose value, so you'll have losses you can realize in order to offset the gains in the index fund. This is not a sound investing strategy -- since capital gains tax is only around 10-15%, depending on your tax bracket, the losses you incur will be much more than the tax savings.

As someone commented, this is what is called "letting the tax tail wag the investing dog". You're buying losers just so you can reduce your taxes, even though it means your net after-tax returns will be lower. This is fine if your goal is to "stick it" to the government, but not if you're trying to save money yourself.

In addition, it's not likely to make that much of a dent in the gains you'll get from selling the index fund. If you've been invested there for many years, your unrealized gains will be sizeable. Meanwhile, you'll only be invested in these individual stocks for a couple of years, so they won't have had a chance to drop much. You're going to invest a small amount in each of a handful of stocks -- even if it drops 10%, that's 10% of a small amount. It's also possible that none of the stocks you pick will have much of a loss at all -- good for your net returns, but bad for your strategy.

Yes, the strategy should work, and I believe is quite a common one these days, with wealth managers offering a packaged version (see https://www.economist.com/finance-and-economics/2025/06/26/how-to-escape-taxes-on-your-stocks).

However, the taxable capital gain is not avoided, only deferred. If you buy a random basket of shares (or make a careful statistical analysis to track the S&P 500 index closely -- the exact choice isn't the issue here) and then sell those which have fallen, you will be left with the ones that have risen, and have a taxable capital gain on those when you want to sell them.

In order for the losing ones to offset the capital gain on your currently held index funds, the size of your investment in the basket of shares must be quite large. You don't say how much you're hoping to offset, but let us assume you'd need to invest at least as much into these shares as you currently have in the index funds. If you already have that cash to hand, might it not be better to keep it safe and use it in a few years' time, so you won't need to sell the index funds at all?

I don’t know why you’re getting such a negative response here. You’re describing something like ‘direct indexing’ which is a legitimate way to enhance ‘tax loss harvesting’. As others have said you have to take care when doing this but besides lots of paperwork I don’t think it’s inherently a risky strategy. To clarify, "random" doesn't play into this strategy at all but as I think you have in mind, you don't actually have to buy all 500 stocks. The question of when to sell and when to buy back is a complex issue with no clear right answer.

There are lots of providers of direct indexing, some with built in execution and others that act as advisors.

united-states

Sort of, but not really. In general, this practice is known as tax-loss harvesting, and can be an effective way to mitigate tax bills by realizing losses on things you already own anyway. Buying things specifically to do this is problematic for two reasons:

• It is not reliable; what happens if you have no losses? Or not enough to achieve your desired offset?
• Your goal in this scheme is to lose money on investments to save money on taxes. The amount you lose will exceed the amount you save, so it is a net loss. It's perfectly reasonable to harvest tax losses on things you already own to mitigate the negative impact of the situation you're in, but planning to do this intentionally is creating a negative scenario to mitigate.

There is a better alternative.

I'm expecting to need to withdraw some money that is currently invested in low cost index funds

Mitigate this need. Your plan is to spend some money over this time on investments that will go up or down, and use losses to mitigate the tax bill on this expected withdrawal. Instead, save that money (cash, HYSA, other low-risk vehicle) to reduce the taxable amount you need to pull from the investment. Sure, you might in the meantime owe a few dollars or cents on the interest these savings accumulate that you would not have owed were it invested elsewhere, but that likely does not offset the tax savings of the taxable gains you realize being smaller.

A more extreme alternative, only mentioned because the proposed plan is "intentionally lose money to minimize taxes", is to quit your job before your do this (assuming you are employed) and/or take other steps to minimize taxable income in the year of the withdrawal. If your only taxable income for the year is long-term capital gains, the first several dozen thousands of dollars (or almost 100k if you are married) are taxed at a rate of 0%. The next half a million or so is only taxed at 15%. This approach would see a significantly lower tax bill than just deducting a couple dozen (or couple hundred) dollars of losses. Of course, if your planned losses are more than that (thousands) it makes way more sense to just save that money to minimize the needed sale rather than do anything more complicated or risky.

While the tax situation is important, you also need to consider the risk that you are already at the peak, and that you may have to sell at a low point if you wait to the last minute to sell the S&P 500 shares.

If you were to start to move funds this year into lower risk investments, and did that over a few years, then the taxes related to the sales would be spread over multiple years. It is true that those funds would not grow very much between now and when you need the funds, but you would reduce the risk that a drop would happen just when you need the money. You could also specify which specific shares to sell to control if these were long term or short term capital gains or even losses.

I voted to close this question as soliciting opinions, but apparently so far I was the only one. So I'll give my opinion.

Generally your approach reminds a mix of direct indexing and tax loss harvesting as mentioned in other answers. But not quite. I personally find it a bit nonsensical. Here's why.

Investment decisions should not be made at random. Even gamblers rationalize their decisions by trying to use statistics and probabilities to their advantage. Generally, if you don't want to think too much about how to invest - the best way is either to devise a strategy that automates the process (e.g.: investing in an index that you believe is covering the areas you want to cover), or hiring someone to do the job (investment manager or investing through actively managed funds), and even then you'll need to do some thinking about the overall strategy.

Randomly picking stock in an index doesn't replicate the index. If you invest large enough amounts in a lot of transactions, the result would be that you converged on an unweighted version of the index, but indexes are usually weighted. However it is unlikely that you'd be investing large enough amounts in significant amount of transactions over time to actually make it statistically plausible. It also doesn't help with re-balancing, which is a necessary part of index tracking (either direct-indexing or through a fund that does that for you). So people who claim that this strategy "mathematically" sound are wrong, it is not.

Tax loss harvesting means re-balancing your portfolio through removing under-performing stocks and replacing them with stocks that you expect to perform better in the future, prioritizing losses. If the thing you're replacing your losses with will under-perform the stocks you sold (because you sold low and bought high, for example), you'll lose money. In the US, wash sale rules are intended to discourage tax loss harvesting (since you can't replace the sold stock with the same or materially similar stock), but it is still a sound strategy in many cases. Just not your case.

Realizing losses is forever. Losing stock in your portfolio is a loss on paper, the stock may rebound and even outperform the rest of your investment candidates in the end. Once you sell the loss - it's yours. So selling decisions should be based on the expectation of the future performance. Tax benefit of loss harvesting comes at the cost of realizing the loss.

Preparing for a large expense. If you know already that in a couple of years you have a large expense and you want to minimize the capital gains exposure - there are other ways to do that. For example:

• Rebalance your portfolio into a mix of more stable investments. Money market funds, HYSAs, treasuries, etc. Your goal would be capital preservation for the expense, not capital investment.
• Gradually start selling over the course of the next couple of years. You'll spread the tax across several years, instead of a single year with a large tax amount.
• If you expect lower income in some years - plan ahead to realize some of the capital gains then to make use of the lower brackets.
• What about leveraging your portfolio and taking an asset-backed loan instead of realizing gains and paying the tax? If you're able to pay it back fast enough, the interest may be lower than the potential tax hit and you'd still keep your investments growing.

germany

If you are not hitting your capital gains tax allowance every year from other sources (dividends, interest etc), it's worth it to sell off just enough to max out that 1k (2k for married couples), if that gain outweighs your transaction costs (which it definitely should).

Apart from that: If you are currently studying or for other reasons drastically lowered your income (such as in retirement): as long as your total annual earnings are below 12.096€ (again, double for married couples), no capital gains tax is required, even if those 12.000 are entirely from capital gains.

To actually work in retirement this does mean that you are limited in how much you can withdraw or earn in total, which can be offset by selling "too much" in one year (and paying the tax) so that the next year you can draw from the cash reserve and only sell for 12k worth of gains (minus other income streams like pensions). Whether that's worth it depends on your details, cash reserves and living standard. Or if you have multiple funds/investments it may be worth "cashing out" the lossy ones strategically to go below the 12k in a year instead of spreading the loss-realization out.

If by "buying random stocks" you mean "mirrorring the index but with individual stocks" then yes, that would be a valid strategy because it allows you to control when losses are realized better as opposed to an ETF where any sale includes the losses and gains from that timeframe.

If you're concerned about the taxes, another option is to leave the money invested and take a loan against it instead. Some investment brokers will give you a loan or a line of credit that's secured using some portion of your investments as collateral. You still own the investments and collect dividends, etc., they just get temporarily partitioned off so that you can't sell them.

The advantage is that a loan is not taxable income, so there aren't capital gains involved. The loan is fully collateralized thus lower risk and lower interest rates. The main downside is that it has to be paid back, but that might not be a problem if you were planning on continuing your contributions to the index fund anyway (just make those into loan payments instead). There is also interest involved, but interest rates tend to be much lower than tax rates.

You'll have to talk to your brokerage for details to see if this would be a good option for your specific needs. It will likely depend on your specific tax bracket, the amount you need to take out, the time horizon on repayment, etc.

Selling your losers means betting that they will not recover. That's committing to losses, which is generally a losing strategy.

Don't sell unless you are going to use the money to buy something you believe will perform better, or you need the money for something other than investments (which should be uncommon if you are investing for a reasonable duration rather than trying to spot trade).

If you are pulling money out of the investments for other uses, the best answer is to do so in a way that maintains your overall investment strategy -- basically as part of a rebalancing action. Typically, that actually means drawing from the investment that is performing best at that moment unless you have reason to believe it will continue that trend relative to the others.

Yes, one can get into loss harvesting if one really wants to. But that means committing to a loss. Don't let the tail wag the dog; don't take losses just so you can claim a loss.