Money? Markets? Autoregressive heteroskedasticity? Ask me about it.
So after being asked by a client why his portfolio was returning way below market averages according to the "Gain/Loss" page in his portfolio, I felt that I had to bring this up. My biggest gripe with these pages is how they compute the cost basis. What do I mean?
Wikipedia has an article on cost basis. Basically, cost basis is "the amount of your investment in property for tax purposes", and it is changes of cost basis that the IRS looks at to assess whether you made money / they should tax you (don't quote me on this). Unfortunately, it also hides a nasty thing — dividends and capital gains reinvested in a portfolio are taxed, BUT also count towards cost basis.
Some math:
Say a stock returned 5% in dividends for a year while its price remains unchanged, and that it declines 5% after issuing the dividends. With a starting investment of $1000, your account after this would be $997.5. So approximately break even (or to be precise, a 0.25% decline). But gain/loss with cost basis is computed by taking the total VALUE of your securities (in this case, $1050) and taking 5% from there - or in other words, a net loss of 5%. To say this is deceptive would be a gross understatement.
The only surefire way I've found to accurately track portfolio performance is the bank transfers in/out approach - take the current account balance, and amortize it with respect to starting and any transferred capital. Unfortunately, most brokers don't track these statistics, so you might have to play around with Excel for a while to do this, especially if you make several transfers (e.g. direct deposit, etc).
End of rant…
