clutch

I live in Toronto, and this bubble is definitely insane, but I can't wrap my head around whether I can bet against it or not. Because to me, the premise "this bubble must pop" is not good enough, obviously because of the timing aspect. In some sense, I need to identify potential catalysts for the bubble to pop, and be sure that the catalyst(s) would happen soon. Here are some scenarios that I though about... - Highly levered, speculative local investors default on mortgage payments or must sell the property for other reasons. This requires one of 1) interest rate rising fast, 2) cannot find tenants (or rental market tanks), or 3) the investors lose their sources of income / wealth. 1) is unlikely given that the government can control it to prevent popping the bubble, 2) is also unlikely looking at Toronto's rental market right now (no sign of population decline) and 3) likely means some sort of global financial melt down, which is hard to predict. - Foreign investors suddenly need to sell, perhaps because they suddenly need the money, or they now have an incentive to sell (better investment opportunity elsewhere or maybe China says bring your money back now and you can legitimately keep it?) or they figure that the investments no longer meets their hurdle rates. I really don't know whether any one can predict timing of these events... Anyone else can think of other catalysts and their estimated timing?

Agreed with others on survivorship bias. Although this is valuable characterization of a potentially successful strategy, we cannot be sure of the probability of success given using this strategy. I suppose you could use Bayes theorem to estimate the probability. I will just use the CoFB population in my estimations. Let S = success (say 10-year annualized return > 20%), X = a particular strategy. We want to compute the probability of success given the strategy, so P(S|X). P(S|X) = P(X|S)*P(S) / P(X) Since OP has found a pretty common strategy among successful investors, we could estimate P(X|S) to be pretty high, let's say 90% or 0.90 (this can be interpreted as the probability that a successful investor has used the particular strategy). Now, what would be P(S), probability of the success? Based on the Jurgis' poll, it was 17%. Obviously, in reality this number would be much lower because not all CoFB members who has 10 year history has voted...Let's say P(S) = 0.10. Given that this is a value investing forum and we have more people who tend to run concentrated portfolios, I'd estimate that the probability of people here using the particular strategy is pretty high. Let's give a conservative estimate, P(X) = 0.30. P(S|X) = (0.90)*(0.10)/(0.30) = 0.30 So even if an investor followed the particular strategy, the probability of the success is still only 30%! And in reality I'd expect this number to be much lower if we took proper sampling of the population.