14 ms·
Wrong. Percentage change is exactly what you look at, if you were to pick just one simple metric.
by solve 12y ago
Wrong. Percentage change is exactly what you look at, if you were to pick just one simple metric.
- joshu 12y agoStandard deviation is the measure of typical amount of change. If you wanted to express it as a normalized z-score, that's your single number. 1% might be a .01 sigma move. It might be a 3 sigma move. Please don't present your opinions on a matter like this as if they are facts, especially when you are ill-informed.
- solve 12y agoOk, sorry my reply was too brief. What I was trying to say is: What would be the most useful single metric that a journalist could use for his audience? Z-score certainly wouldn't be that. I think that "percent change" captures the reader's imagination well, because it rather easily can be converted to an answer to the top question that a reader of a finance article is probably wondering: "What potential returns were there to be made on the market, for this event, had you had knowledge that this event was going to happen?" Otherwise, if you want to answer the reader's possible question of: "How rare is this type of movement?" I would prefer a metric of "mean time between events like this", with "events like this" having some reasonable definition, like versus previous history of this asset, or against similar assets, or against all previous April 1st movements of this asset.
- joshu 12y agoI think the reader's question is: is that a big move? In which case, neither of those really answer the question. In the second you still need to compare the time period to other time periods, or you end up with "is that frequent?" Which is a very similar issue.