There was a recent Warren Buffett article out, which I thoroughly enjoyed, because it broke his returns down into a fairly easily replicable ruleset, which is common with the "academic" definition of value investing plus a few twists. Mainly:
* Value (as measured by P/B)
* Low Beta
* High Quality (as measured by earnings volatility, etc)
http://www.cbsnews.com/8301-505123_162-57524029/how-warren-buffett-beats-the-market/
What I would suggest to synthetically replicat this idea is to run 4 portfolios, each started at a different Q, i.e. P1 starts at Q1, P2 at Q2, P3 at Q3 and P4 at Q4. Then you rebalance each portfolio once a year to the following universe:
* Define the top 25% of value stocks
* Of those value stocks define the top 25% with lowest beta and lowest realised volatility
* Of those value low beta stocks define the top 25% with the highest quality
This is your portfolio for the next year. Apply leverage as desired. You can replicate the Buffett float by shorting corporate bonds if desired.
One important thing to note is there was low/no momentum factor explaining his returns, so stay away from stocks making highs if you want a "Buffett" style portfolio.
Keep in mind, the article points out that Buffetts major difference from other large funds was his reputation, so even in the face of a 44% drawdown, he did not suffer many redemptions and certainly no margin calls. I think that is
the most important lesson. Do not invest in passive equities if you can't handle a 40-60% drawdown. You must be a "strong hand" (with a strong stomach) to successfully execute this kind of strategy. This also means your natural leverage limit is probably somewhat less than 2:1 and more like 1.5:1 at absolute max.
Here is another article that addresses the same topic, lots of pretty charts included:
http://gestaltu.blogspot.com.au/2013/01/track-records-are-rubbish-or-why.html