hi, I'm just reading your new book and struggle with getting from risk weightings to cash weightings (don't call me stupid please ;-)) you have an example somewhere in the middle (i don't have the page number as I read it on my kindle). risk weighting 50/50 for equities and bonds. volatility is 15.2% for equities and 6.1% for bonds. how the heck do you get to the cash weightings of 29.4% and 70.6% ? i know you explain it on page 107 but I don't get it can you explain it please? by the way. I will be in Prague for the QuantExpo. hope to see you there!
Say we had $10,000. If we put $5,000 into stocks and $5,000 into bonds then we'd be getting 15.2% * $5,000 = $760 of vol from stocks and 6.1% * $5,000 = $305. So the proportion of our risk coming from stocks is $706 / (760+305) = 71% and the proportion coming from bonds is 29%. This is completely different from the risk we actually want: 50, 50. So we need to correct for the different volatility. Let's correct both weights so that each asset has the same effective risk as bonds, 6.1% (in fact you can use any number here you like, but it's easier to use a more intuitive number). The weight for bonds is already correct, but for stocks it needs reducing pro-rata, to do this we multiply by the ratio 6.1 / 15.2 = 0.401. The new stock weight is 0.401 * 50% = 20%. So we have: - weight for bonds 50% - adjusted weight for stocks 20% Hang on a sec, though these weights don't add up to 100% - they add up to 70%. No problem, we can just adjust them so they do. To do that we multiply by the ratio 100/70: - weight for bonds 50% * 100/70 = 71% - weight for stocks 20% * 100/70 = 29% Hope that makes more sense now. Yes, hope to see you on Saturday. GAT
Hi GAT, On average about how many contracts in total per day do you trade? Currently, I am trading about 6 contracts per day on a 240K account at 30% volatility across 14 instruments (excluding heavy roll days). Does that seem reasonable to you? Also, I was wondering if you might be updating your historical price csv files in the near future. Would be great to compare the outcome of my own price stitching logic with yours. Thanks.
It probably makes more sense to look at average holding periods or units of risk turned over, as # of contracts is a very crude measure. My average holding period, according to fundseeder, is 30 days which includes rolls. I'll try and find time to do the stitching but I'm rather busy at the moment. GAT
Way better than your peergroup (of sorts), the London megatrendies -- though they're also up this month. Congrats!