http://www.gftforex.com/products/gftmanagedaccounts/history-pac.asp#pachigh Something is wrong with these published returns... Anyone concur?
Yes, I'll say. It is clearly a net losing performance record, incorrectly presented as if it were a net winning record, year by year. Year.....Shown.....Actual 2003.....20.52%.....-0.45% 2004.....14.50%.....-25.65% 2005.....8.00%.....7.79% Nice catch, swoop. This "Pac Tetra High Beta" thing couldn't possibly be registered with the CFTC and/or the NFA? Nah. But, hey, let's look on the bright side: if you speed-read the table title, you'd swear the phrase "high performance" is in there.
In that performance presentation, GFT added up the monthly returns to get the annual returns. What I did was link the monthly returns geometrically. That's the only correct way to compute percent performance over a period of time, given the subperiods' returns. As a simple example, if a hypothetical trader / fund gained 20% in January, gained 20% in February, then lost 40% in March, they would be down 13.6% for the quarter, rather than at break-even: 100 x 1.2 x 1.2 x (1-0.4) = 144 x 0.6 = 86.4 = 100 - 13.6. That's just the way the math works out, for better (when you're winning) or for worse (when you're losing). Ryan Jones and others refer to it as "asymmetrical leverage" (during drawdowns). Incidentally, let's hope that GFT's monthly % numbers are not themselves the arithmetic sums of daily numbers, which would throw them (the monthly numbers) off as well.
Very interesting, when looking at daily and monthly performance for someone to manage your account. Usually different results than what it looks to be at first look. Thanks