The wisdom of the crowd...

[Euler]

I have a false favourite strategy that has been running for, and shows an edge, for twenty years. But of course, that's impossible because the market is efficient.

[ShaunWhite]

I have a false favourite strategy that has been running for, and shows an edge, for twenty years. But of course, that's impossible because the market is efficient.

They say that because they don't understand the word 'efficiency'. Markets are efficient at scale but every individual price is wrong, all the time, everywhere.

(tbh true price is only knowable for a mechanism, not for an animal. So although that's not proveable on individual prices, the idea that everything is right all the time, within the required degree of accuracy, is a statistical impossibility)

[Euler]

I have a false favourite strategy that has been running for, and shows an edge, for twenty years. But of course, that's impossible because the market is efficient.

They say that because they don't understand the word 'efficiency'. Markets are efficient at scale but every individual price is wrong, all the time, everywhere.

(tbh true price is only knowable for a mechanism, not for an animal. So although that's not proveable on individual prices, the idea that everything is right all the time, within the required degree of accuracy, is a statistical impossibility)

I agree

It just annoys me that people keep pumping out how efficient everything is. But they way I see it is that, on average, it is efficient. But that average includes loads of really bad guesses.

[ShaunWhite]

But that average includes loads of really bad guesses.

Hard to escape the omnipresent normalised distribution.

Approximately 68% of data falls within 1 standard deviation of the mean (z-scores of -1 and +1)
Approximately 95% of data falls within 2 standard deviations of the mean (z-scores of -2 and +2)
Approximately 97.5% of data falls within 3 standard deviation of the mean (z-scores of -3 and +3)

The preconception is that a price is the expected probability, whereas it's just the central point of that normalised distribution of probabilities. Without that fact underdogs could never win.

It also explains how a good strategy can have periods of loss and visa versa. Small samples are susceptible to distribution skew, and as sample size increases, mean revision takes over as the dominant pressure. So if you're betting without an edge, the longer you do it the more likely you are to conclude that the market (if not a market) is efficient.

Not news to you, and sounds like teaching grandma to suck eggs, but might be of interest to beginners.

[weemac]

Ox-weight guesses and market prices can be efficient and inefficient at exactly the same time, albeit to differing degrees; the crowd efficiency generally increasing as time progresses while the individual inefficiency remains static.

It seems that most people focus on the numbers and being right or wrong, while seemingly ignoring the passage of time. Time is of fundamental importance in this 'efficiency' argument.

[ShaunWhite]

I agree about time weemac, I see that as the deviation from an efficient price reducing over time as it does in almost all news based markets, while averages remain, average. It's part of the reason BSP is used as a benchmark, it's no better than any other price (cue EMH), but the individual deviations are less than at any other point in time. Scheduled off time is also a point of least deviation.