I am new to the RW Community. It seems to be relatively inactive. Moreover, it is difficult to use. For example, the latest post is not listed at the top of the page, but rather the bottom, forcing one to page down many pages to find it. Also, I have been unsuccessful in downloading .und files posted by other members. It seems that all that I get is a file of html source code for the web page in question rather than the .und file. A new issue of "Tips and Tricks" supposedly is posted every other week, but I have yet to see any new ones posted. Any one have any tips to make this site more user friendly?
“Did you build the backtest extension to 1982 using YQL?”
“How much trouble was that?”
“Any technical problems?”
Once in a while, questioning the Yahoo server (not using YQL) gives no response. I don’t know what they do in such cases, but a few hours before Wall Street opening everything appears to be up and running again. Questioning the server is usually much faster in weekends than during market hours. However, we use the Yahoo server for updating our historical data up to an hour before Friday closing to see how our portfolio selections for the next week is converging.
Like any Dbase, all have errors. As long as they are outliers (including voids) they are easy to find in an automatic and objective way. In our view, such errors appear to be more abundant in dbcmhist than in the various Dbases at Yahoo. Error correction is an area of science, and I know how to do that without distorting or misleading back tests. Of the roughly 6000 US stocks publicly available at Yahoo as US (non-pinksheet) stocks, about 105 need systemic error correction and roughly 25 we exclude as having unreliable data. We calculate the adjusted share prices by ourselves to make sure not to lose significant figures in our back test return calculations. Parsing the fundamental data from the SEC files was somewhat more of a challenge, and errors (outliers) are more difficult to detect.
“So you are able to do, say, $25k+ worth of shares using MOC without problems?”
For our strategy, which only selects stocks with an average trading liquidity in excess of $2.5 million, we are able to place MOC orders not exceeding 1% of that amount. There are people telling us that dividing that percentage by five is better for not inducing price elasticity. All our stocks are non-pink.
“Examining a plot of the rolling cumulative excess return for each of those shows just how quickly and approach can go from adding significant positive value to being neutral or negative value added.”
I think that sounds interesting, but we may be just a bit too practical for that. The market-cap weighted S&P500 had very poor returns during the past 12 years (about zero) with drawdowns that were by far exceeding the risks that we were willing to take. Its returns are significantly different from when you would price or equally weight the portfolios. When you would market-cap weight the ZR#1 portfolios, you practically reproduce the S&P500. Equal weighting and market-cap weighting is like apples and oranges.
“Based on my reference to the S&P website source data, your total return estimates for the S&P traditional & EW indices are off by roughly a factor of 2 (on the hight side).”
Our historical S&P500 curve (brown) exactly follows the Yahoo data for this index. On June 25, 1982, the first date on the curve, the Index value was 109.14. Last Friday it was 1553.28. These two values are the same as reported on the S&P500 site (excluding dividends). Our brown curve (historical S&P500) shows an increase from 1 to 14.23, in exact agreement with the ratio of those two values (1553.28/109.14). Our purple curve shows the random picks of a Monkey out of the universe used, equally weighted, who picks by like throwing a dice with as many sides as the number of stocks in the universe. There are not that many programs available on the market that will show you that curve. When you equally weight those random picks, the Monkey outperforms the equally weighted S&P500 curve by roughly a factor of two over this time span. Throwing a dice allows for multiple picks of the same stock, and each pick, not each stock is equally weighted.
You are lucky it is weekend and I have the time for responding to your detailed questions. But we learn from it, thank you.