Tax Avoidance · Tax Loss Harvesting

Tax Loss Harvesting Update – October 2020

[As with anything tax-related, see my disclaimers.]

As with anything investment-related, see my disclaimers. This post deserves an additional disclaimer. Investing involves risk. I no way do I recommend any course of action. Don't trust random idiots like me on the internet for investing advice. Or for anything else, for that matter.

A couple of years ago, I closed my Wealthfront account. After closing my account, I set out to DIY a tax loss harvesting system, modeled after Wealthfront. I’m tempted to call it my level 5 asset allocation strategy

Building my own tax loss harvesting system has a few advantages, including:

  • Avoid Wealthfront’s .25% management fee
  • Avoid capital gains by donating appreciated shares instead of cash
  • A minor endorphin rush from pushing the buy/sell buttons myself. Yep, I’m a little different

Disadvantages, many of which fall in the “likely overly complicated” category, include:

  • Keeping track of my target allocations, rebalancing as appropriate
  • Preventing wash sales
  • Detecting lots that can be harvested for a loss
  • Not being able to trade as efficiently as Wealthfront’s “robot”

Who Wants To Lose Money?

So first off, no one likes to lose money. But, this is really not a loss. Why? Because after selling an ETF at a loss, I immediately buy another ETF that is highly correlated, but different. Why? To avoid the IRS wash sale rules. Buying a correlated asset means that I remain in the market even after taking a loss, and in theory when the original asset recovers in price, the correlated asset goes up as well. It would only be a permanent loss if I was out of the market and missed out on the recovery.

There are several ways to profit from a tax loss. First, a loss can reduce up to $3,000 of regular W2 income. I’m in the 22% tax bracket, so that’s worth $660 in federal taxes, and another $173 in Virginia state taxes.

If that was the only benefit it wouldn’t be worth it. Why? Because whenever I go to sell it, I’ll possibly run into a capital gains tax of 15% (2020 numbers), which means that $660 of federal tax gets reduced to $450. However, since I regularly give to charity (not virtual signally, I promise!), I gift with appreciated shares that I’ve held for longer than one year instead of cash. Doing this means I avoid paying capital gains taxes on the appreciation. I can use the cash I would have contributed to buy more shares, effectively getting a reset on my cost basis to a higher value.

What Am I Trading?

I mostly adopted the same index ETFs that Wealthfront uses. Just like Wealthfront, I have primary, secondary and tertiary ETFs, and in one case I have a quaternary set of ETFs. Here’s a table, including my target allocations:

Sector Allocation First ETF Second ETF Third ETF Fourth ETF
US Small Cap 9% VXF
Foreign, Small 14% VSS SCZ SCHC  
Emerging Market 14% VWO IEMG SCHE  
Municipal Bonds 5% VTEB MUB TFI  
Resources 5% VDE XLE    


For example, pretend that I have 10 shares of VTI. The price of VTI drops and I harvest a loss by selling the shares. I immediately turn around and buy shares of ITOT with the proceeds. After 31 days, if ITOT falls more, I can sell those shares and immediately buy back into VTI. If instead, ITOT falls before 31 days are up, I can still sell ITOT, but instead of buying VTI and hitting the wash sale rules, I instead buy shares of SCHB. It’s possible to exhaust the hierarchy, in which case I can’t harvest additional loss until the first 31 days expires. Oh well. You can’t win them all

You’ll notice that the US Large Cap and US Small Cap share the primary, secondary and tertiary. That’s because my mix mostly aligns with the internal allocation of those ETFs between large cap and the rest. But if I exhaust my options, I can buy VOO and VXF in equal measure and maintain the correlation. Rarely happens, but it did happen in March.

I do all my trading through my Vanguard brokerage account where ETFs can be traded with zero commission. This system could be done anywhere there is zero-commission ETF trading.

Looking for Losses

My tracking system is a Google Sheet that I’ve built using the GOOGLEFINANCE function to automatically pull the latest share prices. I use the sheet to track both my tax advantaged and taxable assets, and it warns me whenever I risk a wash sale by selling an asset at a loss in my taxable account.

In addition to tracking my gains and losses, my sheet calculates whether my allocation in a particular asset needs to be changed. When do I rebalance? After reading this article by Michael Kitces, I settled on rebalancing an asset whenever it varied more than 20% from its target allocation. 

Here’s my current allocations percentages. It looks like it’s time to rebalance my natural resource allocation. My small cap is also getting close to needing an adjustment.

How often do I check for losses or when a rebalance is needed? Again, after reading Michael Kitces’s article, I settled on checking every week or two. A quick glance at my spreadsheet will tell me if anything is ripe for harvesting, as well as whether a wash sale will happen. It only takes a couple of minutes a week, tops.

October 2020 Update

Nearly 900 words in and I’m finally getting to the point of this article: to give an update on my tax loss harvesting scheme for 2020.

The market has fluctuated wildly this year, mostly due to the pandemic. But even with the wild moves, if I had done nothing, just sitting on the assets that I had at the beginning of the year, to date my total portfolio would have decreased only by -.78%. True, the market dived in early March, but it’s mostly recovered. Far from the best year to invest, but also far from the worst. 

On the other hand, my tax loss harvesting has been a wild success. I had twelve different times that I sold shares to capture losses, all of which happened in the first four months of 2020. Currently, my account balance is 1% bigger than the Do Nothing scenario, if we account for the harvested loss. But despite having mostly the same balance at the end, I also scored nearly $6,000 in losses that can be used to offset my gains from selling covered calls and regular W2 income.

Here’s a table breaking it down by month by month:


Month Do Nothing Gain/Loss Harvest Gain/Loss Harvested Losses
January -3% -6% $54
February -11% -11% $573
March -25% -26% $4,510
April -16% -11% $5,896
May -11% -3% $5,896
June -10% -1% $5,896
July -6% 7% $5,896
August -1% 13% $5,896
September -4% 9% $5,896
October -.78% 12.8% $5,896

And in graph form:

As a side note, 2020 has been another successful year to front load my 401(k), especially with several large buys right during the market dip in March. To date, my 401(k) balance has appreciated 6.83%, most of that attributable to front-loading.


2020 isn’t over. There will probably be additional opportunities to harvest some loss by the hend of the year, depending on the election outcome. We’ll see.

Do you do any tax loss harvesting? I’d love to hear about it in the comments.

Hasta luego!


5 thoughts on “Tax Loss Harvesting Update – October 2020

  1. A few thoughts:
    * You implied that donating appreciated shares required them to be long-term gains. This isn’t a requirement, right? I can still get the tax benefit from short-term gains, right?

    * I’ve read conflicting accounts of VTI vs ITOT as eligible tax-loss-harvesting pairs:

    The much safer approach, I thought, was toggling between the S&P500 index and the Total Stock Market Index.

    Also, I’ve read that index funds are ***slightly*** easier to deal with in implementing the actual trade because the buy and the sell orders are executed simultaneously at NAV at the end of each trading day. With ETF TLH, as you’re doing, I understand that this isn’t feasible. For whatever reason, I just love the simplicity of index funds even though they are almost identical to ETFs.

    (Un)fortunately for me, everything I’ve purchased over the past many years has gone up in price, leaving zero opportunity to TLH. The one exception was during the Covid drop in March where I dropped the ball and didn’t lock in a tens-of-thousands of dollars loss. This was because my money was at Merrill Edge rather than Vanguard and I was reluctant to trade VG funds at Merrill. My bad.

    1. Thanks for the comment!

      • My understanding of donations with short-term capital gains is that the amount you can deduct is limited to the cost basis. You need to hold it longer than a year to get the maximum benefit from the contribution.
      • VTI and ITOT look different enough to me. They are highly correlated, but they follow different underlying indexes. Doing it with the actual index fund instead of ETFs could work as well if not better.

      I’ve come to the conclusion that really you have to be constantly injecting money into a TLH setup to take full advantage of it. Otherwise, the older money tends to continue to rack up capital gains and the only way to get that money generating losses again is to realize a capital gain at some point. At that point, you’re simply arbitraging between your marginal rate and your capital gains rate.

      1. Thanks for teaching me about the LTCG requirement for donating appreciated securities! I had never realized that restriction before; it seems pretty arbitrary, but you’re right.

        VTI vs ITOT are indistinguishable on this plot, particulary for any holding period past 1 month:

        But I agree that it’s a grey area without much legal precedent one way or another. Given the legal ambiguity, if I ever have the opportunity to do:
        1.) Sell total stock market and buy S&P 500
        2.) Hold for 30 days
        3.) Sell S&P 500 and buy total stock market. If stock prices have increased significantly over the past 30 days, perhaps skip this step to not lose some of the TLH benefit.

        Something like March 2020 would have created $100k losses for me in my taxable account had I not been an idiot. That would have created 33.3 years of $3k/year of income offsets. I lament not doing it. In other words, I’m of the opinion that TLH works best for huge market gyrations that occur every decade or so, providing a decade of tax-loss-carryforwards. I wouldn’t worry about the short-term stuff because it seems pretty inconsequential, but perhaps I’m not thinking about it correctly. I guess if an algorithm were doing it, it’d be fine.

        I make the case in the comment section of MMM’s post here that roboadvisors aren’t worth it. Search for “Frugal Professor” in comment section if you’re interested in the argument:

        1. I’m not sure VOO is an less indistinguishable from VTI for more than 1 month. We’ll see if the IRS comes out with more guidance in the future.

          I agree that roboadvisors aren’t worth it. Paying .25% on all of my assets for being able to arbitrage between my marginal tax rate and capital gains just doesn’t make sense. I wrote up my thoughts here, if your interested:

          1. S&P vs total stock market being equivalent in practice is a fine argument, as you demonstrate. For some reason it gives me more piece of mind that the total stock market index has 3500 stocks vs the 500 of the S&P. However, the 3000 firm difference between the two funds seems more defensible to me than tiny differences in how the total stock marked indices are comprised for the different funds.

            In reality, it seems unlikely that the IRS will lay down the hammer for any of this nonsense anyway. They are incredibly understaffed and outmatched when it comes to enforcing this type of minutia.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.