Calculating Portfolio Returns




Late last year, my broker (TD Direct Investing) finally introduced online performance tracking – something that was sorely lacking. After all, investors should know what their portfolio performance has been.  If you can’t measure how well your portfolio is performing, how do you know whether or not you are investing wisely.  Are you investments doing better than just leaving the money in the bank (hopefully the answer is yes!)? More importantly, is your portfolio beating the benchmarks and or equivalent index funds that you could be investing in.  If not, it may be worth re-visiting your investment strategy or going back to index investing and other simpler forms.

When I blogged about my Year in Review last year, I had a lot of difficulty extracting performance data.  Being a little better prepared for this year’s review (soon to be arriving), I’ve been generating data points to help me better my understanding of performance; while I wait for the information online to improve.  Last year, all that I had available was available was book value and market value of investments held at that moment in time.  How is that a problem?  Let’s take for example:

Demo Portfolio

Holding Book Value Market Value
Google (GOOG) $100K $150K
Apple (AAPL) $100K $105K
Tesla (TSLA) $100K $200K
Cash $100K $100K
Portfolio Total $400K $555K

In an example like this, TD Direct Investing would basically say I made $155K profit – good for a gain/loss of 51.67%.  Perhaps the one good calculation that TD does, and this is debatable, is it ignores the cash component of book value – after all, if you didn’t invest it, why include it in the return calculation?

Besides that, there are a couple of other issues:

  1. Doesn’t account for past transactions – i.e. if stocks were bought and sold, the profit (or loss) get lost in cash or in the subsequent investments made as we only get to see what is currently in the portfolio
  2. Doesn’t account for cash inflow or outflow – the return displayed is really just market value over book value of each individual stock; not the portfolio overall.  Doesn’t include any deposits made, withdrawals made, or any dividends received.
  3. This doesn’t account for any time.  51% return (or 39% if you include cash) looks amazing – who wouldn’t want a 51% return?  But what happens if this portfolio is as a result of investing for  5 years, 10 years, 20 years, or even 30 years?  That 51% return really works out to become annual rates of 8.7%, 4.3%, 2.1%, and 1.4%

Working Backwards – Solving Issue #3

No surprise here – using the Compound Annual Growth Rate quickly resolves our time problem.  In our example above, the 5 year return looks pretty decent, but beyond that, we might look for better investments over those same time frames.  Speaking of time frames, I absolutely hate it when people get focused on calendar year returns – they’re great reminders to do an annual check-up on your portfolio, but in reality, who is really buying on January 1st, and selling on December 31st?!  The price of any stock, mutual fund, bond, etc. is bound to move up and down throughout the year.    It just makes it easier to determine return from a market value vs. book value perspective.  Ultimately, what I would eventually like to do is track my individual stock buy and sell performance against a benchmark index or group of indices – that way, you can tell how each investment decision has performed vs. the benchmark, and aggregate that across the portfolio.

But back to CAGR – here’s where it doesn’t work so well.  Let’s say it’s 2010, and your neighbour, Bull Picker, tells you about this company called Universal Display (OLED) – maker of next generation LED technology.  Fast forward to 2015 and you go out to Best Buy and see all these OLED TVs and remember your conversation with Bull.  You go home and check out the stock price for OLED:

Source: Google Finance

Source: Google Finance

Yowsers – it’s gone up from $11 all the way up to $36 today – a clear 2-bagger and good for a CAGR of 27%!  Next time you see Bull, you say, “Man, I wish I had followed your advice on OLED – you must be making a KILLING!”.  Bull throws up his hands, curses at the sky, and says “That piece of crap?!  I’m barely breaking even.”  Turns out Bull has been putting in money throughout the 5 years at some pretty inopportune times:

Date Purchased Buy Price Shares Profit/Loss
19-Feb-10 $10.95 100 $2505
15-Oct-10 $26.33 100 $967
13-May-11 $45.08 100 -$908
7-Oct-11 $44.31 100 -$831
27-Apr-12 $45.16 100 -$916
16-Apr-12 $23.06 100 $1294
6-Sept-13 $35.88 100 $12
20-Dec-13 $33.92 100 $208
5-Sept-14 $36.91 100 -$91
Total  $30,160 $2240

This is really problem 1 and 2……irregular cash inflows and outflows make it difficult to determine the true return.  Ironically, it’s also why mutual fund performance is also a little deceptive:

Source: TD Canada Trust

Source: TD Canada Trust

It basically assumes you made your deposit 1 year, 2 year, 3 year, 5 year, or 10 years ago on January 31 and left it there without depositing or withdrawing any further funds.  It’s also a little unclear if the historical performance includes the re-invested dividends as in the chart.

So then, of course, the fund companies need to provide a calendar view in the event that you did make annual contributions:

Source: TD Canada Trust

Source: TD Canada Trust

Of course, I can’t fault mutual fund companies for reporting this – after all, they do need to provide some record of the fund’s performance over time.

A Simple Solution (XIRR)

For individual investors like me, who don’t have portfolio management tools, we turn simply to Excel.  Built directly into Excel is the XIRR function – which is basically an IRR calculator that accepts irregular cash flows.  As my reference to the article in Money Sense indicates, you basically provide dates and amounts of:

  • starting amount or deposit
  • cash inflows (deposits)
  • cash outflows (withdrawals)
  • ending amount (current market value)

and let Excel do the rest.  The result is an annualized return over the start to end period.   It’s a simple and quick formula that will probably satisfy 99% of the performance metrics you want to figure out.  At the end of the day, looking at something portfolio wide, we really only care about how much money we put in, how much we withdrew, and the total market value today.  That market value today not only includes stock prices, but also incorporates any dividends received whether they are received in cash or re-invested.

Interested in your own multi-timeframe performance (i.e. 3 month, 9 month, 1 year, 2 year, etc.)? Just find the starting value at the beginning of the period and XIRR the transactions during the timeframe, and the end amount.  Within a year, you can then compare that rate with any benchmark index you want or a series of benchmarks.  However, where the Money Sense article says take your XIRR value and compare it against the 1 year return of the index (i.e. S&P 500 gained 12%), I like to run a parallel calculation as if the deposits and withdrawals I made was in the benchmarks.   I’ve started to use XIRR in my dividend portfolio updates and we’ll see more use of it in my year-end review (coming up soon).

The most important drawback I’ve noticed – over short timeframes, the number becomes extremely exaggerated (due to the result coming back as an annualized rate).  Let’s say you were able to catch lighting in a bottle, and you experienced a 10% return in just a month.  XIRR would basically assume that you could consistently obtain that return – giving you an annual return of 207% which while amazingly well, probably isn’t achievable.  Depending on how much that stock accounts for the overall portfolio, that large bump might be felt across many months.

Ultimately, the most accurate and widely used performance method is the Time-Weighted Rate of Return – and while I would love to use it, requires me to capture market values for every cash inflow and outflow which is a) not readily downloadable and b) probably not worth the effort for the additional accuracy.

Given the difficulty in data, I made the effort to try to find ways to determine portfolio performance.  However, the end goal would be to actually measure performance on an individual buy or sell transaction – something that we’ll leave for the years to come!


Disclaimer: The author owns shares of GOOG, AAPL, and use to own shares of TSLA. The author can also relate to Bull Picker in the unfortunately timing purchases of OLED. A full list of holdings can be found here.


2 thoughts on “Calculating Portfolio Returns

  1. Pingback: Portfolio Year In Review – 2014 | Fearless Cal's Investment Journal

  2. Pingback: Portfolio Update – January 2018 (+8.4%, Currency -1.7%) | Fearless Cal's Investment Journal

Leave a Reply

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

You are commenting using your 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