Tuesday, July 13, 2010

Measuring Portfolio Returns

On the surface, measuring your portfolio’s return and comparing it to some index seems like it should be easy: take the final value, divide by the starting value, and compare. However, there can be a number of complications. This is my attempt to evaluate my investing performance.

Over the years I’ve done many approximate calculations to assess my stock-picking ability. A few months ago I met with Preet Banerjee of Where Does All My Money Go? fame, and he asked me how my stock picking results compared to index returns. This innocent question left me stammering because I knew I had only approximated the answer. Without taking into account all factors, I couldn’t be sure.

I fired my financial advisors and started investing on my own roughly the middle of 1998. I completed a nearly full transition to index investing roughly the middle of 2010. So, the period of interest for measuring my stock-picking results is this span of 12 years.

However, I did not start with one lump sum of money. I made literally hundreds of deposits and withdrawals to and from 9 different trading accounts. Figuring out a rate of return over these 12 years requires an Internal Rate of Return (IRR) calculation. (There are cases where IRR does not work well, but that is not the case here as it turns out.)

There were two different ways to go with this calculation:

1. Track all the account withdrawals and deposits.

2. Track all the stock-related transactions such as trades, commissions, fees, dividends, and income tax payments.

I decided to go with approach 2 because I want to focus on stock-picking results and not mix in results from cash, bonds, and other non-stock investments. Here are some statistics from the large spreadsheet I generated:

– 144 months
– 88 stock purchases
– 91 stock sales
– 398 dividend payments
– Many commissions and fees that I lumped together each month

Also included on the spreadsheet were all the income taxes I paid on capital gains. I did not include taxes paid on dividends because these taxes would have to be paid on index dividends as well. This gives an apples-to-apples comparison between my return and index returns when the index returns include dividends but not income taxes.

A minor quibble is that the index approach would have a bigger built-in capital gain (or at least a smaller capital loss). It’s hard to value the accumulated capital loss my portfolio generated that I am able to carry forward. This difference puts my return at a disadvantage compared to indexing, but I’m just going to ignore it because the combination of available RRSP and TFSA room makes it unlikely that I will ever use up all of this accumulated capital loss.

(Some readers may wonder how a fairly successful portfolio could build up a large capital loss. The answer is related to technical differences in the ways that stock gains and stock option gains are taxed.)

I performed the IRR calculation and also calculated the stock index returns for Canada and the US for the period from 1998 July 1 to 2010 July 1:

Me: 9.6% per year (my portfolio accounting for capital gains taxes)
Canada: 5.9% per year (S&P TSX Composite total return)
US: -1.6% per year (S&P 500 total return measured in Canadian dollars)

So, I beat the index by a decent margin. Is this edge statistically significant? Am I some sort of stock-picking savant? Hardly. Let’s try a little sensitivity analysis.

For many years I held stock in the company I worked for and this continued long after I stopped working for them. I obtained most of this stock through stock options. So as not to throw off the calculations, I treated the options as though I purchased them at the value they had when I was first allowed to exercise them. Over the years I exercised the options and sold the resulting stock in bits and pieces. At the end, the remaining shares were nearly worthless.

What happens to my returns if we eliminate my 3 biggest stock sales one at a time? I recalculated my compound average return assuming that these trades never happened and that I held the shares until the same day I got rid of all the other remaining nearly worthless shares. Here goes:

Return without biggest sale: 2.6% per year
Return without 2 biggest sales: -1.0% per year
Return without 3 biggest sales: -2.4% per year

My results on this stock completely swamped all other stock trades I made. The employer shares I sold early gave phenomenal gains and the ones I held to the end gave dreadful losses.

In the end I don’t really know if I’m any good at stock picking, but I doubt it. I’m content with my decision to switch to mostly indexing, and I’m happy to have been lucky for 12 years.

13 comments:

  1. Excellent post. All investors would do well to carry out this exercise. One thing that makes your situation difficult is that you have 9 trading accounts. I have found that if you can combine accounts it makes it easier.
    Also, if you take the full value of your accounts when there are meaningful cash flows you can see if asset allocation is adding value.
    I think you came to the right conclusion when you concluded that indexing was the way to go. Like you I outperformed but I was never convinced it was skill. I in fact came close to buying Enron but pulled back at the last minute! At one time I had held Fannie Mae but sold it in 2002. So it is worthwhile to replay your thought process - it's not just about the bottom line number. In fact, I think some mutual fund managers get caught up in their stock picking prowess because they are lucky over a given period only to show dissapointing results down the road.

    ReplyDelete
  2. @Passive Income Earner: I agree that the 15% discount should not count towards your investment performance. I think it makes sense to include any company stock or options in your ROI as long as you treat them as having been purchased for fair market value at the first time you are allowed to sell them.

    @DIY Investor: Sadly, I can't combine most of my trading accounts. I will be able to eliminate the two I had for my children, but the remaining 7 can't be combined for tax reasons. This number will go back up to 9 when my wife and I each open a TFSA.

    ReplyDelete
    Replies
    1. The first reply above is to Passive Income Earner's comment:

      Thanks for the post. I should look at how I am doing against an index.

      I get a 15% discount on our company stock every 6 months through the ESPP plan but I don't tend to include those in my investments and ROI. Same for my options and RSU. They are not present anywhere in my net worth calculation because of the vesting timeline and risk of a zero value. Since I don't pay for them, I don't include them.

      Delete
  3. Great post! The first few years I started investing on my own were horrific. Two big losses put my overall returns into the negative. I think most people put too much emphasis on their wins and discount their losses thinking "I won't make a mistake like that again", only to go on to other mistakes.

    ReplyDelete
  4. @Preet: Thanks. One other thing I noticed in this exercise is that it would be very easy to ignore some bad results. Like any accounting work, if you want to get a certain result, you can rationalize certain choices to push the final answer in a given direction. I had to slap myself a couple of times and stick with a realistic view.

    ReplyDelete
  5. Cry Havoc and let loose the Dogs of yet another Can't Lose Stock Picking Methodology!!!

    I am sure you could write a book about how well you did, without espousing using the method you used and still make money on the deal. Then you can write a follow up story about future successes that didn't follow your old system, stating that you had rethought your ideas, and make more money, etc., etc., etc.,

    ReplyDelete
  6. @Big Cajun Man: My best advice is to be lucky. Perhaps an infinite improbability drive would help. Failing that, just go for the index. I guess that makes for a very short book.

    ReplyDelete
  7. @Saint Lucian Dutch Canadian: I wish you luck (and maybe some skill). Curiously, I was more concerned about money and risk when I was young. That seems to go against the conventional wisdom. It's not that I'm a big risk-taker now, but I'm not as stressed about potential losses as I was when I was young. I think the explanation is that I understand more about investing now.

    ReplyDelete
    Replies
    1. The comment above is a reply to Saint Lucian Dutch Canadian's comment:

      I'd say you did pretty well. For my rrsp (15.5% of my gross) I used td e-series index funds but I have a little bit of money that I use to pick stocks in a non-reg acct with the hope that I can do well. I guess I just feel like I'm young and invincible.

      Delete
  8. @CC: Ah yes! Incubation is one of the keys to mutual fund success (at least for the management company). I guess if I just throw investor money into the index for 12 years and charge 2% MER, my average excess return over 24 years will still be +1%!

    ReplyDelete
    Replies
    1. The comment above is a reply to Canadian Capitalist's comment:

      I think you should start a mutual fund. Your "incubator fund" beat the market by close to 4 percentage points over 12 years!

      Delete
  9. My method was to set up an Excel sheet and every time I make a purchase, I simulate an equivalent purchase in the S&P500 index. Over time, I can compare how I did compared to an identical S&P500 purchase (my recent trading has been in US stocks).

    I don't give myself credit for dividends received, and allow the S&P500 to accrue a 1.5% dividend. This simplifies my calculations and seems pretty conservative.

    ReplyDelete
  10. @Gene: I thought about doing my calculation by adding or subtracting the appropriate number of index units. I decided that this might be harder, but I agree that it is a good approach. If your dividends tend to be 1.5% less than the S&P 500 dividends, then you'll get a good answer. My guess is that your dividends would be higher and so you're being somewhat conservative.

    One part I found tricky was accounting for taxes.

    ReplyDelete