Executive Summary
Capital loss harvesting has long been a staple of investment tax strategy - so much that the Internal Revenue Code has special "wash sale" rules to ensure that the technique is not overly abused. Fortunately, though, the wash sale rules can be navigated effectively, allowing taxpayers some means to take advantage of available tax losses.
However, while tax loss harvesting remains a viable strategy, it is often greatly overvalued, as the true benefit is not the tax savings from harvesting a loss but merely the benefit of deferring those gains. In the meantime, the strategy has a non-trivial exposure to several risks, including the potential for the alternative investment held during the 30-day wash rule period underperforming the original investment, the possibility of negative tax arbitrage if the investment rebounds in the near term, and the danger that harvesting losses too effectively over time will drive the client's future capital gains into a higher tax bracket! In addition, the fact remains that capital loss harvesting produces no benefits for clients who are eligible for 0% capital gains tax rates, and in fact potentially harms them; in such scenarios, clients should actually be harvesting gains, not losses!
Ultimately, this doesn't mean that harvesting capital losses is a bad strategy, but it is a strategy where the risks must be carefully considered, as they can easily outweigh the relatively modest benefits!
Tax Deferral, Not Tax Savings
One of the fundamental points of confusion about capital loss harvesting is that it doesn't permanently save taxes, it merely saves them temporarily - in essence, it just defers them. The reason is that when capital losses are harvested, the new investment position after loss harvesting has a new, lower cost basis - which means if the investment is ever intended to be sold or consumed during life, someday the taxes will still be due.
For example, assume an investment was purchased for $100,000, but is now worth only $90,000 due to a significant market decline. The investor harvests the $10,000 capital loss, which generates a $10,000 loss deduction resulting in $1,500 of current tax savings at a 15% tax rate. However, after harvesting the loss, the investment only has a cost basis of $90,000, which means if the investment ever recovers to $100,000 again, there will be a $10,000 gain, resulting in a $1,500 tax liability at the same 15% tax rate. Thus, in the end the $1,500 tax savings now is offset by a $1,500 tax increase in the future, and the net result is $0 of tax savings (which makes sense, since the investment started at $100,000 and ended at $100,000)!
Of course, the caveat is that the tax deduction occurs now, and the tax liability for the recovery occurs in the future. Thus, the true value of harvesting the capital loss is the opportunity to invest the near-term tax savings for growth. For instance, if the $1,500 tax savings (which was 1.67% of the original investment, or $1,500 / $90,000) can be invested at an 8% growth rate, then the investor can earn a $120/year of economic value. However, relative to an investment that was worth $90,000, this true economic value of loss harvesting is a remarkably small benefit of only about 13 basis points (0.13%) per year! And of course, the benefits must be weighed against the potential costs to complete the transaction; with a $90,000 investment it might be appealing, but with a smaller investment position like one worth only $9,000 (and therefore an economic value of only $12/year of tax dollar) the raw trading cost to execute the sales and (re-)purchases may exceed the entire benefit!
In addition, there are also important indirect costs and risks to consider...
Tracking Error Risk During Wash Sale Period
One challenge to harvesting losses is that under the "wash sale" rules, the client who sells an investment for a loss must allocate the money elsewhere for 30 days to avoid having the loss disallowed. If a "substantially similar" investment under IRC Section 1091(a) is purchased during the interim time period, the loss cannot be claimed, and instead is added to the cost basis of the replacement investment (which means the loss is ultimately deferred to the future, although if the security is repurchased inside an IRA the loss may be forfeited permanently!). The wash sale rules can be avoided by investing in something else that is similar - but not "substantially" similar - or by simply keeping the money in cash or something else completely different.
Of course, if the goal is to remain invested, taking the money out of the market and into cash or a completely different investment is unappealing. In fact, owning anything that is materially different than the original investment may be unappealing - after all, if the investor wanted to actually own something different, the investment could have just been sold outright in the first place. The whole point of "loss harvesting" is to capture the loss without materially changing the investment!
As a result, investors try to buy something that is as similar as possible to the original investment, without running afoul of the rules. Historically, this was done by buying another stock in a similar industry - for instance, harvesting a loss in Ford and buying GM during the intervening time period, or harvesting a Merck loss and buying Pfizer instead. In today's world of pooled investment vehicles, the lines have blurred a bit, and it's less clear about what exactly constitutes an investment that is not "substantially similar" - selling a large-cap index and buying a small-cap index would almost certainly be fine, but it's less clear in the case of selling a large-cap index like the S&P 500 and buying a total market index instead, given that the returns of both will be dominated by the same set of (cap-weighted) stocks.
The reason why this matters is that if the investment results are too different - even just for a period of 30 days - there's a risk that the entire 13 basis points of economic value for harvesting the loss will be potentially lost by return differences as the temporary new investment fails to track the original investment effectively (not to mention merely any potential drag due to possible differences in expense ratio). Of course, the goal of many loss harvesting investors is to purchase an investment similar enough to reduce any exposure to significant tracking error. Yet the reality is that if the investments track too perfectly (for instance, mutual funds or ETFs that mimic the same underlying index), it runs afoul of the wash sale rules!
In other words, avoiding the "substantially similar" wash sale rules essentially requires the investor to be exposed to a material amount of tracking error! Of course, there's also a possibility that the replacement investment could outperform rather than underperform - not all "tracking error" is bad! - yet nonetheless the reality is that almost by definition, the investor can only harvest a loss if the replacement investment places the entire value of loss harvesting at risk to a comparable or greater amount of tracking error!
The Risk Of Negative Tax Arbitrage
In addition to the risk of tracking error overwhelming the economic value of harvesting a loss, a secondary challenge is the risk that the investment rises in value during the 30-day period, effectively converting a long-term capital loss into a short-term capital gain.
For instance, if an investment purchased 2 years ago for $100,000 has fallen to $90,000, and the $10,000 long-term capital loss is harvested, but during the interim period the investment rebounds to $98,000 and the client wants to switch back to the original investment, the rebound will trigger an $8,000 short-term capital gain. Fortunately, the loss can offset the gain if they fall in the same tax year, but notably the capital gain/loss ordering rules require long-term losses to offset long-term gains first. As a result, if the client already had $10,000 of other long-term capital gains, the loss harvesting transaction effectively converts a $10,000 long-term capital gain into an $8,000 short-term capital gain. If the client is subject to a 25% ordinary income tax rate (along with a 15% long-term capital gains tax rate), the client actually turned a $1,500 tax liability (15% of $10,000 long-term gain) into a $2,000 tax liability (25% on $8,000 of short-term gain)! This is essentially a form of negative tax arbitrage - the investments don't materially change, but the tax rate moves in the wrong direction! Of course, as with the tracking error scenario, it's possible that the investment won't change in value, or that it will go down even further and that there will be an additional short-term loss to harvest. Nonetheless, the scenario represents another form of risk to the loss harvesting transaction - and with volatile investments, a potentially material risk given how small the benefits of loss harvesting are in the first place! In addition, there's always the risk that within a year after the wash sale is completed, the investment will be sold after rising in value, triggering a(nother) short-term gain on the original investment that would have been long-term if the loss was never harvested in the first place!
An even greater form of negative tax arbitrage emerges from the fact that in today's tax environment, there are effectively four long-term capital gains tax brackets. As a result, if an investor systematically harvests $10,000 of losses from a portfolio every year - reducing cost basis by $10,000/year in the process - after a decade of the strategy a subsequent liquidation could trigger a whopping $100,000 capital gain in a single year (or much more when accounting for not just recovery back to the original purchase price, but real growth on top of that). If the loss harvesting occurs enough times over enough years, it creates the potential of such large capital gains in the future that the client ends out in a higher capital gains tax bracket down the road! Thus, if the capital losses are harvested at a 15% tax rate, but they create so much more gain in the future that eventually the recovery is taxed at 18.8% (with the new 3.8% Medicare surtax) or even at 23.8%, the client could destroy far more value by boosting the tax rate than was ever gained in benefit from the loss harvesting strategy itself! And the negative tax arbitrage is even worse if the losses are accidentally harvested for clients in the bottom two tax brackets - who are actually eligible for 0% long-term capital gains rates, and should actually be harvesting gains instead!
Of course, if investments are ultimately held until death, they receive a step-up in basis, which is a form of positive tax arbitrage (as the gains are effectively at 0% but the losses were harvested at some more favorable rate). Nonetheless, if the client ever intends to sell the investment to spend during life - and/or ever anticipates making an investment change - then the reality is that a step-up in basis is not really an option, but the risk of negative tax arbitrage remains a danger.
Ultimately, none of this means that capital loss harvesting doesn't have any value - there is still some benefit, and the larger the loss (or the higher the tax rate) the greater the value (although the recent proposal to require average cost accounting for all investments would partially limit the ability to harvest losses strategically). But the reality is that unless the tax rate is very high or the loss is very large, the benefit is actually quite limited, as it is based not on the amount of tax savings, but only on the economic value of deferring those taxes to the future! And in the meantime, the risks remain; in fact, some risks like exposure to tracking error are essentially a requirement to avoid the wash sale rule in the first place, and while larger losses create a greater harvesting benefit they also create a greater exposure to negative tax arbitrage and higher rates in the future! And that's before accounting for the raw transaction costs of loss harvesting in the first place, which may be fairly low in today's low-cost trading environment, but still represents an important threshold when considering how small of a loss is worth harvesting at all. In the end, all of this doesn't take loss harvesting off the table altogether... but it does mean the (limited) benefits must be weighed carefully against the costs and risks inherent in the strategy.
