Ann: Half Year Results Analyst Presentation, page-45

Just some basic rules of thumb that I codified and wrote down and follow strictly - if goes 10% past what I feel the value is (based on my model) then I put it on a watching brief. If it goes 30% above what I think the value is then its in my sell zone and if it goes 50% above what I think the value for the next year then its a mandatory sell (that's what happened here). Basic tenet is that businesses revert to their value but that that doesn't stop businesses continuing to "pop" when they are on a run so that's why I let it run to 50% above fair value.

My back test with things I follow (higher quality businesses with strong free cash flow and a decent return on equity) is that basically, every 10% extra over value means a corresponding 20% - 25% increase in the risk of it falling back to fair value within the next 6-12 months.

I also use 80% trailing stop losses of the 52 week high - generally works well since I don't have invest in micro caps. That's just based on the mathematics of a 20% fall requiring the stock to go up 25% from that point to get back to "even" and that's more a capital preservation thing for when things I buy trend down even below what I think is fair value - so the basic idea there is to cut the losers.

Let winners run (to a point), cut losers. Sounds simple and I let the rules above, and a few others, guide me and take the emotion out of it.