A big selling point of the automated investing industry is that it gives Main Street investors low-cost access to sophisticated strategies once limited to wealthy clients of Wall Street banks. The technique increasingly mass pitched to ordinary savers: using computer-driven tax moves to eke out extra profits.
Selling losing stocks to generate losses that can offset winners, or tax-loss harvesting, is marketed as a daily ritual by robo-advisors from Schwab to Wealthfront. An algorithm scans an investor's portfolio each day in search of shares that have declined and sells them. The losses are used to offset taxable investment gains or other income. Unused losses never lapse: Up to $3,000 — per transaction, so cumulatively, a multiple of that — can be rolled over for reducing taxes on gains in future years.
The more frequently losses are intentionally taken, the bigger the opportunity to scrape out some extra profit in a retirement portfolio, the industry's pitch suggests. The technique is now cast as akin to brushing your teeth daily (technically, having your teeth brushed for you), not cracking open a fine bottle of wine.
Not so fast, according to recent research.
Roni Israelov, the president and CIO of NDVR, an investment advisory firm in Boston, and Jason Lu, a research economist in the economic modeling division of the IMF, sought to answer the question of when tax-loss harvesting is worth it. The tactic is increasingly popular amid the rise of low-cost exchange-traded funds and direct indexing, which allow investors to create custom benchmarks. They wrote in a
Among their findings: There's a concrete threshold at which the strategy pays off. There are also time frames that work better than others. Not harvesting frequently enough can leave money on the table. But a program that's too aggressive can leave an investor underweight in individual stocks, potentially locking her out of gains or causing her portfolio to lag whatever benchmark it may be trying to mirror.
"Harvesting on a daily frequency offers virtually no advantages to harvesting monthly," they wrote. The paper, published on SSRN, contributes to a growing body of research that's critical of the strategy.
The strategy is widely promoted, but it can leave some investors worse off.
Amid the market volatility that has been 2022, tax-loss harvesting is touted as an opportunity.
"This scenario is tailor made for strategies like TME," said Rajesh Nakadi, the head of investments – global family office for BNY Mellon Wealth Management. "Tax-managed equity" is the bank's term for tax-loss harvesting.
Until robo-advisors automated tax-loss harvesting in recent years, the strategy was a time-intensive, cumbersome process involving lots of documents for accountants and the IRS, making it the purview of the affluent. But as ETFs, which seek to mirror gauges such as the S&P 500, and index funds, which allow investors to create custom benchmarks, have burgeoned, they've become a focal point for the technique.
Thresholds
The Israelov-Lu paper focused on harvesting losses from individual stocks, not from ETFs. Unlike mutual funds, those trade on exchanges like stocks.
The paper found that for a stock comprising 0.5% or less of a portfolio, it doesn't make economic sense to harvest losses monthly unless the security is down at least 10%.
For daily harvesting, the paper found that a stock needs to be down at least 15%. Any less and the investor typically loses money — the one-off tax savings are eroded by the cost of buying back the stock, according to the paper's simulations of returns by 1,500 securities over 20 years. The thresholds also apply to investors harvesting losses in a stock that's a bigger chunk of their portfolio, but only if they use a complicated "throttling" strategy to toggle between being overweight and underweight in the stock over time.
"Do portfolios benefit by taking advantage of daily harvesting opportunities?" the paper asked. "Perhaps a counterintuitive result, but our research suggests the answer is 'no,' at least when applied to individual equity positions."
The thresholds apply only to harvesting in which the same stock is repurchased. Under the IRS's wash sale rule, designed to prevent gaming of the system by selling a stock to book a tax loss while immediately snapping it back up, an investor has to wait 30 days before repurchasing the same share or buying something "substantially identical" (by contrast, a so-called similar stock can be bought immediately, such as Occidental in place of ExxonMobil.)
The paper casts tax-loss harvesting as essentially a form of risk arbitrage, in which investors game out the reward of banking usable losses against the risk of skewing a portfolio and missing out on larger gains.
The 30-day lag can force an investor to take on the added risk of potentially missing out on a stock's quick rebound and put him in the undesirable situation of selling low and buying high — the opposite of what he wants. Or the repurchased stock may rebound so much that the taxes owed when it's eventually sold outstrip the tax savings from harvesting earlier.
Or the entire exercise may be pointless. Say you bought $10,000 of Facebook stock years back and it's now worth half that. Harvesting the loss and waiting 30 days to buy back the stock translates into a $5,000 capital loss, the equivalent of a $1,190 tax benefit for someone paying the long-term capital gains rate of 23.8%, which includes the Affordable Care Act levy. Now say the shares you bought back for $5,000 climb back to $10,000. When sold, the capital gains tax due will be $1,190 (=.238 * 5,000) = the same as the benefit from harvesting.
Despite what the wealth management industry suggests, some investors benefit more from the popular strategy than others.
Things get trickier with stocks that pay dividends. If you haven't owned a dividend-paying stock for at least 60 days and it pays out — typically once every three months — the dividend doesn't get the lower qualified dividend rate, which is the same as the long-term capital gains rate. In that scenario, the investor owes ordinary rates, now a top 37%, plus sometimes an additional 3.8% for the net investment income levy. Using the strategy means that "there's a little bit of market timing involved," said Andrew Hambleton, a financial life advisor at Telemus Capital in Chicago.
Robo-advisors
With Wall Street having an awful year after a long bull run that began in 2009 — stocks are down roughly 16% so far this year — the strategy is increasingly prominent with automated investing platforms.
Schwab Intelligent Portfolios, a robo-advisor,
The risks aren't generally conveyed to or understood by ordinary investors, according to Luke Smith, a co-founder, partner and chartered financial analyst at AlphaWorks, a quantitative equity portfolio management and research firm in Jersey City, New Jersey. He added that any fees for the service "are likely to overwhelm the benefit."
The Israelov-Lu paper didn't set out to address the behavioral financial implications of thresholds at which to harvest, noting only that the wash sale rule "places loss harvesting into a classic risk-reward paradigm, creating a trade-off between harvesting yield and active risk."
But for investors whose advisors hand-pick losses to harvest or do so on their own, volatile markets can tap into deep-rooted biases that can prompt unwise decisions.
"Tax-loss harvesting makes you look at the markets at precisely the time when, behaviorally speaking, maybe you shouldn't be looking at the markets," James M. Dahle, an emergency room physician and the founder of the White Coat Investor, a blog for medical professionals,
