After reading E. Chan's Algorithmic Trading and also reading the Kalman Filter post I wanted to try to replicate the EWA/EWC trade from 2006 to 2012. For reference, here's one whose returns are similar to Chan's. Unfortunately, I've been having a very hard time replicating the returns of Chan here on Quantconnect. If you simply swap the tickers and time range to EWA/EWC and 4/26/2006-4/9/2012 you get a disappointing result (attached). However, the most interesting thing is that the pricing of EWA/EWC is much different than the data says. 

Looking on Yahoo and QC's own data (EWC, EWA), the prices are much higher than what are traded at in the backtest (clone and run or look at the trade download here). The backtest prices are ~$11 and $18 vs ~$22 and $25. 

Any ideas to interpret this? Am I looking at this wrong? Does this have anything to do with it?