Not Just Another Article About Investment Tax Management
Tax loss harvesting has become one of the most widely sold ideas in wealth management. And why would it not be? I have met exactly zero people who enjoy paying taxes. But I want to explain why I think it works in theory a good deal better than it works in practice. Specifically, I would like to share why I believe an investor who understands the whole marketplace of strategies (or who hires someone who does) keeps more of what they earn than one who sees only a slice of it, and why independent oversight across one or several institutions may add value that has nothing to do with performance or fees.
Tax Loss Harvesting – A Modern History
By 2009 (a few years into my career at that point), tax loss harvesting was not much of a focus. The world had gone from seeking return at any cost to protecting capital at any cost, and losses were plentiful; nobody needed to manufacture them. Then came the zero interest rate years, and the gains of that period slowly but surely consumed the realized losses many investors were carrying from 2007 to 2009. By 2014 systematic tax loss harvesting was back in vogue, generally dominated by two providers. Those providers were followed in short order by newer entrants offering substantially the same strategy, with better technology, at a lower price. A good analogy here would be comparing this to the way the cloud followed the mainframe.
What is being sold
The general pitch is this: for a fee of roughly 0.10% to 0.35%, a manager will track an index of your choosing, the S&P 500 or the Russell 3000 or MSCI EAFE, while selling constituents that decline past a set threshold and replacing them with similar exposures.1
The logic is simple and rather elegant. In a typical year a substantial minority of the stocks in a broad index finish lower even when the index itself finishes higher. A stock that has declined from its cost basis produces no value for you (beyond any dividends received) until you realize the loss by selling it. And when you do, you have created a tax asset that can offset a realized gain of the same character, now or in the future.
The providers describe the benefit as tax alpha, which most of them put at 1% to 2% a year for equity strategies.2 That figure is a manager’s estimate, and it assumes every dollar of realized loss is put to work against a gain of the same character, short term against short term above all. That is the first of many ifs.
What tax alpha actually is
It helps to be precise about where the value comes from, because the answer is more interesting than the sales material suggests.
Every harvested loss lowers the cost basis of what you own by the same amount. The tax you saved this year is, for the most part, tax you may owe later when the replacement position is sold. Harvesting is mostly deferral, which is not a small thing, and it is not the full story either. The tax you did not pay stays invested and compounds until the deferred gain is finally realized, which over a long enough horizon is real money. Realized losses may pull a taxpayer down a bracket in a high-income year, so the offset can be worth more than the headline rate suggests. And where a position is held until death, the potential for a step-up in basis turns deferral into elimination. Each of those is a genuine source of value, and each depends entirely on your own circumstances, your gains, your tax rates, your time horizon and your estate plan, which is precisely the point of this piece.
Where it goes wrong
These strategies are now everywhere. I have met a number of investors who own two or three of these tax loss harvesting strategies at different wealth managers and private banks. In each case, the investor was sold on the idea in good faith by someone who never took the time to truly understand if the tax benefits could be fully realized. In more than one case the programs were quietly cancelling each other out, as were the active managers overseeing portfolios of single stocks.
The reason is the wash sale rule, which disallows a loss if you acquire substantially identical securities within thirty days before or after the sale.3 The rule applies to you as a taxpayer, not to any one account or custodian, so a purchase at one firm can disallow a loss taken at another. It reaches your spouse’s accounts on a joint return.4 And under a 2008 revenue ruling it reaches your IRA: a purchase there within the window disallows the loss in the taxable account, and because the IRA’s basis is not increased, the loss is gone for good rather than deferred.5
Positive intent, negative outcome. Your CPA is the right person to walk through your own situation; my job is to make sure the programs across your accounts are not working against one another in the first place. By the time it makes it to your CPA, it is too late.
Beyond long-only
So far I have been describing the managed equity universe. Similar strategies exist in fixed income, where they earn their keep in rising rate environments, and in long/short equity, which brings us to hedge funds.
Please allow me a paragraph before the pitchforks come out.
The category is full of misunderstanding, thin investor education, headline failures and fees that look high relative to the value people perceive. I spent four years running hedge fund sales and strategy for the United States and Canada at a global private bank, working with some of the wealthiest families and family offices in the world and the managers in which they sought to invest. I have some sympathy for the skepticism, some pride in my “uncommon knowledge” of this space, and some experience of what the hedge fund structure actually allows for.
What makes these vehicles interesting for our purposes is that they hold long and short positions, and the research on that structure is now quite clear. A long-only direct indexing program as discussed above, absent new contributions, tends to top out with cumulative net losses somewhere near 30% of the capital originally invested. It takes about eight years to get there but is ultimately dependent on market direction (i.e. larger declines closer to initial investment will generally accelerate tax losses harvested). A tax-aware long/short strategy run with an appropriate amount of leverage can pass through net losses near 100% of invested capital within a few years.6
These strategies come in two forms and the difference matters. In a separately managed account you own the stocks yourself, reporting arrives on a 1099, the minimums run into the tens of millions of dollars through the large platforms, and because the positions are yours the wash sale rule applies across all of your accounts; the manager coordinates around its own trades and no one else’s (or within a very limited scope). In a commingled fund you own a partnership interest, the fund’s own trading sits outside your personal wash sale footprint, minimums are a fraction of the SMA, you receive monthly capital account statements showing the tax character of what has been realized to date, and the K-1 arrives by March. Long-only direct indexing, for comparison, is available from about $250,000.7
Portfolio real estate
Consider the arithmetic. If a long-only harvesting program occupies 25% of your portfolio and produces 1% of tax alpha on that sleeve, you have earned 0.25% across the whole. A long/short tax aware sleeve that is a fraction of that size can produce a similar result at the portfolio level. What matters is what the small sleeve leaves free. Three quarters of your equity allocation is not spoken for by a tax program. The sleeve can be added to, trimmed, or moved between managers with different risk profiles as conditions change, and it is far easier to keep feeding a small allocation than to keep adding to a 25% position in United States large-cap stocks, where most investors are already overweight. That last point is not academic. Every harvesting program, long-only or otherwise, draws its losses from the basis of the money you put in. Stop adding capital and a day arrives when the embedded gains are too large for any more losses to be found.
Losses, of course, are only worth the value of the gains they offset. Like everything else in portfolio strategy, this lives in a balance rather than in an absolute.
Platform decisions
I have spent a good part of my career building and implementing tax-managed/aware strategies for taxable investors inside large institutions, including a long/short strategy that families and family offices across the country still use today. The experience taught me something about how availability works. Whether a given strategy is offered to you at a given firm is a platform decision. It turns on capacity, on minimums, on whether the firm can achieve scale with it, and on the priorities of a diligence team with a long queue. It is made without reference to your particular circumstances. That is simply the reality of how a large financial institution must operate to achieve scale and minimize risks for the institution, not necessarily for you as an investor.
What it means for you is that the absence of a strategy from a menu tells you nothing about whether the strategy works, and the language used to explain the absence, which tends to run toward the reassuring space of a moral high ground, tells you even less. A strategy can go on performing, and investors can go on benefiting from it, only if they know it exists. An adviser who is not tied to any one platform can inform you and educate you about what other investment opportunities exist in the overall marketplace.
What the coordinated version looks like
Imagine the alternative. We start by reading your last several years of tax returns and identify where losses were disallowed and what that cost you. Portfolios holding actively managed equities or ETF portfolios across two or more managers, even at the same financial institution, are the ones most at risk. We then seek to improve the tax efficiency of how the whole portfolio is run, including: holding assets with more punitive tax implications in tax-deferred or tax-free accounts, benchmarking after-tax returns across the total portfolio, and the timing and the methods by which distributions are taken, to name a few. We use the industry’s latest investment management, wealth, retirement, trust and estate planning tools to make sure you own the right assets, in the right account types, titled the right way, at the right institutions, coordinating with your CPA and your attorney(s) along the way as needed.
Those steps can add meaningful value, and we have yet to say a word about performance or fees.
This is why I believe the wealth management industry will keep moving toward an independent model for families with $5 million or more. The largest institutions devote their most precious resources to who they perceive to be their most precious clients. Then they look to implement for scale elsewhere, which is why your harvesting program sits in its own dedicated account rather than being woven through the management of your whole portfolio. True investment tax management, as opposed to being tax aware, is the whole portfolio.
If this piece left you a little dizzy, that is an honest reaction. It is as complex as it sounds. It is also what I do for a living, and I do it as a fiduciary, which means the decisions are made in your interest as the law defines it, rather than in someone’s perception of it. Want to learn more about how this may apply to your specific set of circumstances? Conversations are unhurried, confidential, and without obligation. Simply reply to this email and I will find a time that works for you.
1 Parametric Portfolio Associates LLC, Form ADV Part 2A dated April 21, 2026, Item 5 (Custom Core Equity 35 basis points, Tax-Harvest Core 20 basis points; standard schedules, negotiable); J.P. Morgan Tax-Smart SMA platform, 23 to 10 basis points by account size, as reported at launch (InvestmentNews, April 2022).
2 Parametric, “Active Tax Loss Harvesting in Fixed Income: Checking In at Midyear,” July 2025, citing third-party research of 1% to 2% for equity and 0.3% for fixed income.
3 Internal Revenue Code § 1091(a).
4 IRS Publication 550, Investment Income and Expenses, “Wash Sales.”
5 Rev. Rul. 2008-5, 2008-3 I.R.B. (January 22, 2008).
6 Liberman, Krasner, Sosner and Freitas, “Beyond Direct Indexing: Dynamic Direct Long-Short Investing,” Journal of Beta Investment Strategies, Fall 2023; Sosner et al., “Loss Harvesting or Gain Deferral?” AQR working paper, September 2023. Historical simulations, not the results of any product.
7 Parametric Form ADV Part 2A, April 21, 2026, Item 5 account minimums.
