Theoretically, shouldn't any indicator-based or relatively simplistic strategy have no edge at all?

[R1234]

There are tons of weak edges out there. The vast majority of people look at the weak edges and toss them out maybe because the sharpe is 0.5 or has losing years, etc. But interesting things happen when you combine non-correlated weak edges. The sum will be greater than the parts.

 

[Pekelo]

There is a faulty premise in the OP, namely that one can only make money by having an "edge."

Being long in a rally or short in a selloff is not an edge, yet profitable. Also different people have different timeframes so just because someone is selling that doesn't necessary mean they are shorting the market, they may just getting out of a much earlier long.

TL;DR: No.

 

[lindq]

If you're using an indicator, or some combination of indicators that any algorithm could easily backtest, wouldn't algorithms have arbitrated any edge away?

If there exists a pattern that has an edge, wouldn't an algorithm be able to trade it far faster and earlier than you? Quants could easily get paid to write an algo that tries out every combination of indicators on every timeframe...

Granted, by the same logic, that edge would eventually no longer exist to the algorithm either thru alpha decay...

Suffice it to say, it seems that any good edge is outside the realm of inside-the-box thinking with regards to indicators or simple price-action.

I'm a fan of technical analysis, but if I follow the logic of "if it works, somebody or some machine has already leveraged it to the maximum" then it seems that only leaves fundamental predictions?

Or is it like Heisenberg's uncertainty principle; everybody interacting with the same pattern will fundamentally alter the pattern so that it doesn't end up being what everyone predicted, and we can only see the true patterns in hindsight?

There are thousands of instruments that can be traded, in an infinite number of time frames and combinations.

I wouldn't lose any sleep over your secret sauce being stolen, and wouldn't let that concern keep you from trading.

 

[guowei58]

If you're using an indicator, or some combination of indicators that any algorithm could easily backtest, wouldn't algorithms have arbitrated any edge away?

If there exists a pattern that has an edge, wouldn't an algorithm be able to trade it far faster and earlier than you? Quants could easily get paid to write an algo that tries out every combination of indicators on every timeframe...

Granted, by the same logic, that edge would eventually no longer exist to the algorithm either thru alpha decay...

Suffice it to say, it seems that any good edge is outside the realm of inside-the-box thinking with regards to indicators or simple price-action.

I'm a fan of technical analysis, but if I follow the logic of "if it works, somebody or some machine has already leveraged it to the maximum" then it seems that only leaves fundamental predictions?

Or is it like Heisenberg's uncertainty principle; everybody interacting with the same pattern will fundamentally alter the pattern so that it doesn't end up being what everyone predicted, and we can only see the true patterns in hindsight?

Instead of guessing if you have an edge or not, just trade the strategies that are working so far and get rid of them when they stop working for whatever reason...and pick them up again when they start working again...it's not an "intelligent" approach but sure beats getting into a theoretical discussion that doesn't have good answers.

 

[Blitzjoker]

There is a faulty premise in the OP, namely that one can only make money by having an "edge."

Being long in a rally or short in a selloff is not an edge, yet profitable. Also different people have different timeframes so just because someone is selling that doesn't necessary mean they are shorting the market, they may just getting out of a much earlier long.

TL;DR: No.

I would have thought that a trading strategy leading to that outcome is exactly what an edge is.

 

[Pekelo]

I would have thought that a trading strategy leading to that outcome is exactly what an edge is.

So we have a definition problem here. For me, an edge is some kind of advantage over other participants in the market. Being long in a rally is not such thing, so not an edge.

If you define edge as being profitable, then why not just call it being profitable?

TL;DR: Traders/investors can make money without an edge.

 

[Blitzjoker]

So we have a definition problem here. For me, an edge is some kind of advantage over other participants in the market. Being long in a rally is not such thing, so not an edge.

If you define edge as being profitable, then why not just call it being profitable?

TL;DR: Traders/investors can make money without an edge.

Ok; so you can have an edge without being profitable by your definition? To me, having an edge is having some kind of systematic approach that enables you to make gains in the market. And of course you can be profitable without having an edge (for a while anyway) just by good fortune so there is a difference.

So we can agree to differ on that one, but I think my definition is more useful. By your definition you could just lose money more slowly than the average person, so actually not participating in the market at all would be an 'edge' wouldn't it, as the majority of people lose money in the markets?

 

[tomorton]

So we have a definition problem here. For me, an edge is some kind of advantage over other participants in the market. Being long in a rally is not such thing, so not an edge.

If you define edge as being profitable, then why not just call it being profitable?

TL;DR: Traders/investors can make money without an edge.

Surely edge means an edge over random, its not defined by the other participants.

Please, I don't want to have to climb into the loft to get my Mark Douglas's out...

That said, yes, many experiments have shown that random entries can be so well managed as to be profitable. In that case the money management would be the edge.

 

[Pekelo]

We actually agree.

1.Ok; so you can have an edge without being profitable by your definition?
2.To me, having an edge is having some kind of systematic approach that enables you to make gains in the market.
3. And of course you can be profitable without having an edge (for a while anyway) just by good fortune so there is a difference.

1. That is not what I said.
2. Agree, but that doesn't mean all profitable approach is an edge itself.
3. See? We agree.

 

[Pekelo]

Surely edge means an edge over random, its not defined by the other participants.

I never said random. Being long in a generally bullish market environment is not random.