Anybody know why selling out of the money options is a way to game the Sharpe Ratio? At first I took this this as a no brainer b/c selling these are ways for a portfolio to generate extra income, however I saw this again last night and started thinking about it in more depth. Sharpe Ratio is a measure of return per unit of risk. Extra income from the sale is a part of that return, and there is risk in the transaction that the option could be assigned (this may be very low, but it is still there. Otherwise there would be no buyer on the other end). so looking at the transaction this way I see no reason not to include these in the Sharpe Ratio. The only time I see this as a gaming of the ratio would be during a transition of performance periods. If a mgr were to sell a European call that does not expire until after the end of the performance measurement period the portfolio would benefit from the extra income without the risk of being assigned. However, the material does not call out this specific case and only uses the general case of selling out of the money options. Does anyone else agree that this gaming is relavent only in more specific cases than what they propose?
Well, in your case the option expires in the measurement period, so risk would be captured. I believe they are referring to an option that expires after the reporting period.
I’m not in the mood to argue with CFAI right about now. If CFAI tells me my name is “Goat”, I’m gonna say yes. The arguments can start right after I get those three letters stamped after my name.
I really didn’t mean to ask - Who wants to gang up with me against the CFAI… I’m really wondering why they indicate this gaming as a general case. The material takes alot of caution to be as specific as possible in most areas. In this case they do not mention any specifics or go into greater detail on the subject, which leads me to believe that there must be a deliberate reason as to why. What I meant to ask is if anyone knows why the general case would be considered a gaming of the ratio.
me.tega Wrote: ------------------------------------------------------- > I’m not in the mood to argue with CFAI right about > now. If CFAI tells me my name is “Goat”, I’m gonna > say yes. The arguments can start right after I get > those three letters stamped after my name. So, once you get the charter your card will read: Goat, CFA
I think it is related to normality assumption of Sharpe you will sell otm options and generate small profits (does not matter when they expire, simply compare premium against MV or final settlement of the options) and it will be many small profits. The standard deviation of the result from this option trading does not show the potential of the risk when one option goes bad and you lose X times more than you made on the other options. Now I am thinking how you measure % return when you sell options…
bpdulog Wrote: ------------------------------------------------------- > Well, in your case the option expires in the > measurement period, so risk would be captured. I > believe they are referring to an option that > expires after the reporting period. This is what I have understood based on CFAI readings. Options are also not normally distributed, but in this particular instance the timing issue was the reason they gave. I think non-normality would also be a factor personally, over a longer var measurement period in which the option would expire (or whenever if it is American), but they did not state so.
so you say if option expires after reporting period, its market value is not included in portfolio return calculation, only the premium? I dont think so.
I love these lists we gots to remember
pfcfaataf Wrote: ------------------------------------------------------- > so you say if option expires after reporting > period, its market value is not included in > portfolio return calculation, only the premium? > > I dont think so. First off I just said thats how CFAI presented it. You can solve it the correct way on the exam, let me know how that works out. Second, if you write a call and collect a premium and then your reporting period ends two days later, how much do you think the market has moved in most situations? VAR won’t capture the skewed returns distribution (negative due to unlimited downside and only the premium upside) until a longer measurement period captures the volatility of the option’s market value.
markCFAIL is correct. The key to the gaming part is that you collect income now on long tailed risk be it long term options or other risks that won’t show up on your current period risk measurement.
wake2000 Wrote: ------------------------------------------------------- > I love these lists we gots to remember you mean “goats” to remember… in line with the rest of this particular thread…
Ok, so I looked back into this again tonight and saw that they are reffereing to the fact that all the risks may not make themselves present through variability of returns for “several years.” So this seems like a time frame issue to me. for example: if you measure a sharpe ratio over a given year, you may not capture the entire risk in option writing position. I can accept that…until I think of other assets in the portfolio that risk may not be fully reflected over the course of a year. If we think of a growth stock whose variation of returns may not be fully captured until the completion of a full business cycle we can see that measuring the sharpe ratio over the course of just a year may seriously change its value. Again, I’m not trying to argue with the CFAI, I just think there are other issues to think about here. It makes me think that due to stationarity issues that sharpe ratios cannot be compared from one period to the next. Just a thought.
Let’s look at this thing from another perspective. If you buy traditional assets like stocks, you can only dream of making over 100% once in a blue moon and losing 100% on investment is also at the tail end of the curve. That is not the same with options. If you buy a call option the chances of making over 100% returns are much higher and at the same time the maximum you can lose the whole premium is also higher. This makes the investment positively skewed. Using this kind of options games the Sharpe ratio. Selling an out of the money options also makes the investment negatively skewed as you can possibly lose more than 100% of investment in a heart beat while your maximum gain is 100%. This also makes Sharpe ratio looks dumb. I hope I get extra points for arguing in favor of CFAI.
Options change the return distribution. Rather than approximating a normal distribution, options produce skewed, kurtotic, or leptokurtotic return distributions, depending on the choice of option types and strikes. The standard approach to estimating volatility is unbiased even if the underlying distribution is not normal, as long as the distribution has finite variance, so in theory options using options should not bias the estimate of volatility. However, since the Sharpe ratio is generally calculated using relatively few data points, strategies that ‘sell the tails’, such as writing out-of-the-money puts or calls, can often survive several years without being hit. For example, writing a 10% out of the money put on a portfolio indexed to the SP500 each month would probably generate 2-2.5% in annual premiums. Based on the empirical distribution of monthly returns, this strategy has a 2/3 chance of surviving three years without paying off once, and a 50% chance of surviving five years. If the manager is lucky, this strategy will show a significantly higher Sharpe ratio, as the premiums flow directly to the bottom line with no apparent increase in volatility. Strategies that involve taking on default risk, liquidity risk, or other forms of catastrophe risk have the same ability to report an upwardly biased Sharpe ratio (witness the Sharpe ratios of ‘market-neutral’ hedge funds before and after the 1998 liquidity crisis). source: http://www.hedgeworld.com/research/download/howtogameyoursharperatio.pdf
what page of cfai book does it mention gaming the sharpe using options?
p88, luck number. …alternatives: Sortino ratio; Gain-to-loss ratio; Calmar ratio; Sterling ratio.