April 6th, 2009
A dedicated reader commented yesterday concerning the SSO, the Proshares Ultralong S&P 500:
“These things only track daily moves (even then they are spotty), not moves over a period of time.
If you invest $100 and you lose 10% in one day, you need 11% the next day to get even. A 10% loss in an ultralong fund would lose $20 of that investment, and would need 25% (a 12.5% up day) the next day to get even.
A move up of 11% on that $80 still leaves you with a loss since you are tracking daily moves and not compounding.
I hope I’m not misunderstanding your premise, but the ultra’s just don’t work over time unless you string together similar movements day after day.”
Sorry, I’m humbly disagree. Last June, just as the Market was about to nosedive, I wrote a piece entitled “How Do Ultrashort ETFs work?” in which I included some x-y charts with regression lines. Notwithstanding the weak opening this morning, it’s appropriate since the market seems to be bottoming to answer the readers question about using the Ultralong ETF by including a graph of it’s long-term behavior:
When you enlarge the chart, you’ll think that I incorrectly marked Friday’s closing values for the SSO at 31.00 (the actual close was 22.12). The difference is the cummulative dividends paid by Proshares on the SSO since its inception:
The SSO without adjusting for cumulative dividends would have the following relationship with the S&P 500:You can see the imperfect relationship between the SSO and the S&P 500, starkly different than the relationship when cumulative dividends are added back. Adding the cumulative dividends back are appropriate if you are a long-term holder of the SSO. If you’re a very short-term trader, the SSO may be subject to errors the readers mentioned.
But, moving forward from today, if you bought the SSO and held for an extended market run (several months or years and 500 points or more) then you’d evaluate the position with all future dividends added in. If the market carried the S&P 500 to 1000, an 18.7% increase, the SSO should increase by 45% 32 (or 45 on a dividend adjusted basis).
If my analysis is wrong, please don’t hesitate telling all of us how.