Thursday, March 29, 2012

Comparing My Personal Yearly Returns to a Benchmark

Stock pickers should work out their personal investing returns and compare them to an appropriate benchmark. Otherwise you have no idea how well you’re performing. On the other hand, ignorance is bliss, except for the part where you have to keep working past retirement age.

I’ve evaluated my past performance in many ways, but rarely have I done it on a year-to-year basis for my entire portfolio. After much fighting with a spreadsheet and my old account statements, I have personal return figures for each year from 1995 to 2011. These returns take into account all my investments, capital gains, dividends, incomes taxes, mutual fund redemption fees, transfer fees, margin interest, administrative fees, trading costs, and other account fees.

A particular challenge was how to handle employee stock options. I decided to treat them as though they didn’t exist until the first day I was permitted to exercise them. On this day I recorded a purchase equal to the current value of the options. So, it was as though I bought the options on that day. Any gains that occurred before this day would not count toward my investing prowess.

Return figures don’t have much meaning unless we compare them to something. For this I used index returns from a Libra Investments spreadsheet. Each year I figured out what percentages of my investments were in each of Canadian bonds, Canadian stocks, U.S. stocks, and foreign stocks. I used these weightings to compute a blend of the Libra figures for long-term Canadian Bonds, the TSX Composite, the (U.S.) Wilshire 5000, and the MSCI EAFE. This blended figure is then my benchmark for that year.

And the results are (drum roll, please) …


1999

Obviously, the first thing that leaps out is the wild return from 1999. This was the result of an insane bet with almost everything I had on one stock. I was extremely fortunate to have this work out well for me. I would never do anything this crazy again.

1995 to 1998

This is roughly the period of time where I used a couple of financial advisors. The results actually look pretty good until you remove the effect of some split corporations that I owned for several months in 1998. These were similar to leveraged ETFs that invested in Bell Canada Enterprises (BCE) and telecom companies. These bets worked out very well. Once you remove their gains from the 1998 results, the advisor returns trailed the benchmark. Overall, my advisors lost to the benchmark by about 2% per year.

1999 to 2009

This is roughly the period where I was a stock-picker. 1999 was spectacular, but after that, my results lagged the benchmark by a little over 1% per year, on average. For an analysis that isolates my stock-picking results excluding index ETFs and advisor investments, see this earlier post.

2010-2011

I started buying nontrivial amounts of index ETFs in 2009 and the percentage of my portfolio that was indexed increased steadily to over 90% today (my only individual stock now is Berkshire Hathaway). 2010 was particularly painful for my stock-picking results. The part of my portfolio invested in indexes had a return of 22.6%, but once you include my fabulous picks, my return for 2010 was 12.9%. Ouch.

Overall

The compound average returns from 1995 to 2011 (inclusive) were as follows.

My returns: 9.6%
Benchmark: 5.5%

So, I outperformed by 4.1% per year. However, I can show that this was blind luck. If the stock I made the huge bet on had traded during 1999 at the same price it cost by the end of 2000, my compound average return would have been only 1.2%, and I would have lagged the benchmark by 4.3% per year. Essentially, I rolled the dice and won big.

Conclusion

For several years I had some fun feeling like a big wheel talking about which stocks I liked and didn’t like, but now I’m happy to throw away all my old annual reports and focus on other things. I believe that switching to index investing will make me richer in the future. At some point I expect that I’ll sell even my beloved Berkshire Hathaway stock, but not just yet.

6 comments:

  1. Wow very nice returns, especially in 99 lol.

    ReplyDelete
  2. Ok, spill it Michael. What did you buy in '99?

    ReplyDelete
  3. What a refreshing mixture of humility and mathematical rigor. If only all DIY investors had that combination.

    It's fascinating that your huge 4.1% "alpha" is almost entirely the result of one amazing year over a decade ago. While the mutual funds never have extreme returns like that, there are many examples of fund managers who had a great run a long time ago that will make their overall results look good for a long time. But, of course, they tell you absolutely nothing about what to expect going forward.

    Great post.

    ReplyDelete
  4. @Miiockm and @Patrick: You could check historical returns for stocks that tripled in 1999, but I'm afraid that wouldn't narrow it down much :-)
    The end of the tech boom was crazy.

    @Dan: Thanks. I try not to lie to myself. I fear that too many stock pickers delude themselves the same way that lottery players delude themselves into thinking they're ahead of the game.

    ReplyDelete
  5. Quite interesting. Thanks for that.

    I think you could reasonably add your stock options, where the "purchase price" would be FMV when the options vested.

    ReplyDelete
  6. @Blitzer68: That is essentially what I've done, if I understand your meaning. I included stock options using a purchase price at FMV on the first day I was permitted to exercise them and sell the resulting stock. Vesting is one barrier, but most companies place further restrictions on when employees can trade company stock (e.g., right before earnings are announced).

    ReplyDelete