Portfolio Loss Harvesting Part 2: Trying to Mimic the S&P500

While one of my goals is to recognize tax losses from my portfolio, I have to be careful not to sacrifice pre-tax gains in achieving that goal.  The way I try to meet this goal is to come up with a representative sample of stocks that tracks the S&P500 relatively closely without actually having to purchase all 500 stocks that are part of the S&P500.

Before going any further, I should mention that I’m almost a complete amateur when it comes to investment.  I know a few things about the market and I know a few statistical and financial concepts.  But my background is in tax.  Therefore, any apparent success (or failure) I have in implementing my pre-tax strategy may well be random and not the product of any real business acumen or lack thereof.  Someone who’s really savvy about this stuff may well poop on my strategy, but that’s okay.

In any event, as I said, my plan is not to buy all of the components of the S&P500.  Instead, I start off with a smaller pool of stocks that I can invest in that would, nevertheless, hopefully track the S&P500.  The idea here is similar to presidential polling.  You know the ones—where they poll 1,000 voters and determine how millions of people will vote.  With this in mind, I’d also expect a small sampling of stocks would also track well (if not perfectly) against the S&P500.

Very generally, I pick stocks with large capitalization and high liquidity.   I had a hunch that these would be the least volatile and the least costly to buy and sell (e.g., low bid-ask spread cost) which, for an amateur, is probably the safer thing to do.*

In acquiring the stocks, I tried to split them across sectors (e.g., industrials, materials, information technology, etc.) in a way that mimics the S&P500 sector split.  I must admit that my eagerness to run this strategy caused me to have some very big mismatches which I’m trying to fix over time.  For example, I have way more exposure to Utilities than the S&P500, which causes my other buckets to be slightly off as well.  As indicated, I’m actually working on fixing this as I make new investments or sell off old ones.

Based on the above, I currently own shares in 55 companies.  In addition to these, I have sold shares in 35 other companies.  In the aggregate, I’ve dealt with about 90 different companies.

Now, since my goal was to track the S&P500, I periodically check to see if the stocks that I buy and sell are, in fact, tracking the S&P500.  I do this in a couple of ways.  First, I create a portfolio on Google with my stock (including the ones that I have sold) and I map them against the S&P500 backdated to when I started this project.  I then use my eyeball to see if the lines are close.**  As you can see below, my portfolio does track relatively closely to the S&P500.

pool versus spy

There are some flaws in this chart, however.  First, the chart presumes I bought all my shares at the beginning of the strategy whereas I am continually purchasing and selling shares.  Therefore, it doesn’t capture the changing nature of the investment.  Second, I have to include stocks that I no longer own.  Why?  Well, I only keep winners; that is, my tax strategy is to dump the losers and recognize the loss.   Matching my winners versus the S&P500 would make me look like the smartest investor of all-time.

And that brings me to the second way I determine how successfully my pre-tax strategy is working.  I do a side-by-side comparison of how my actual strategy is doing compared to how a hypothetical investment in SPY shares would have worked out for me.  I do this for several reasons.  By doing this, I can reflect purchases and sales over time.  Moreover, I’m viewing SPY as my alternate form of investment. That is, if I wasn’t doing this I could just dump all my money into SPY shares and be done with it.  Therefore, I want to make sure my pre-tax strategy doesn’t stray too far away from how I would perform if I simply owned SPY shares.  I don’t expect it to exactly match SPY performance but I don’t want to be too far adrift, particularly in the downward direction.  So, onto the results.

Well, before I give you the results, I’m going to give you a caveat.  As you can see from the chart above, I started off my strategy at a relatively low point in the last few months.  Not the lowest point, but a fairly low one.  Therefore, the yields I’m going to set forth are going to seem otherworldly.  This is also exacerbated by the short time period over which the price change has occurred.  Therefore, you should look at these results only from a comparative point-of-view.  That said, the pre-tax yield my strategy has been about 48.76% through today.  The pre-tax yield on a hypothetical investment in SPY shares would have been about 36.22%.

Okay, that is obviously good in that I’m ahead of my hypothetical alternative investment even before adding in the tax strategy that I’ve implemented.  However, it might also indicate something problematic in that my performance may be too far ahead of SPY, suggesting that I’m not tracking it as closely as I thought.  Some potential explanations for this deviation:

  • since I’m not doing simultaneous trades, I can never get exactly the “right” price I would have paid for SPY when I buy stock X. Because I do my actual trade first and then see what price I would have gotten for a SPY share next, any potential lag would cause some mismatch.  If prices are generally going up when I buy, then I would expect the SPY performance to lag.
  • I noted earlier that my sector mix is not exactly in line with the S&P500 mix. This could explain much of the mismatch.  Also, as noted earlier, I’m working on fixing this.

Or, perhaps it’s too early to draw any conclusion because the actual dollar value of the difference is not that large and the yield calculation is simply exaggerating the effect due to the short time frame.

That said, I can’t say that I’m unhappy about being on the positive side of the variance (for now).

loki smiles

Up next, I’ll discuss triggering losses for tax purposes and how I go about making those decisions.

*Note, in a later post, I will discuss how this may not necessarily give me the best result from a tax perspective where I will also discuss bid-ask spread.

**I’m trying to re-learn some of my statistics so I can actually determine mathematically how close it really is rather than eyeballing it.  For now, I’m eyeballing it.

2 thoughts on “Portfolio Loss Harvesting Part 2: Trying to Mimic the S&P500

  1. Tom,

    Hope things are going well. Lebow pointed me to your blog. You should check out wealthfront — they have some very smart guys doing this strategy quantitatively — not familiar enough with their retail offering to know whether it makes sense.

    Also, makes more sense to talk about your returns in absolute percentage space (as opposed to annualizing the returns).

    Would be great to catch up some time.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s