4 ms·
There is a subtle error in the way this article is phrased. Since the outcomes are mutually exclusive you can sum them when computing the expectation. And you'r
by carlob 13y ago
There is a subtle error in the way this article is phrased. Since the outcomes are mutually exclusive you can sum them when computing the expectation. And you're also neglecting all the high probability-low gain outcomes.
- boxerbk 13y agoThe high probability-low gain situations would actually lower the value of the options. This model was very generous to the right-skewed data. Low gain situations will typically be worthless because of VC liquidation preferences. Also, these acquisitions are typically unannounced, so it's impossible to guess their value. Can you explain the error with summing? I did sum the expected value of an acquisition and an IPO. Is that wrong? I'd like to fix it for others.
- carlob 13y agoIn general the expectation of making X is given by \int x dP(x). In simpler terms suppose you have a 9% chance of making $1000 a 1% chance of making $10k and a 90% chance of making $100, your expectation is .09*1000+.01*10000+.9*100 == 280 In you article you only take into account a fraction of the total probability that is equivalent to assuming that in the vast majority of cases you'll make exactly zero. That would mean lowering the previous estimate to $190.
- boxerbk 13y agoUnderstood. My assumption was that in the vast majority of cases, you would make zero because either the company fails or the exit is so small that VC liquidation preferences would negate any common stock. You are right, it isn't clear in the article. Thanks!