This question was asked to me some years ago in a newspaper, and I took the risk to give an absolute, "No! Never!"
Then, Darwinex started to take off, and I suddendly realized that algo traders are now everywhere in the retail side, EVERYWHERE! And some developers - real genius - are now all the top traders in the hall of fame. At this point, I was about to abdicate and change my mind.
But here comes a 3rd phase: the one where $THA and $VTJ are capped at 3MM AuM or so, and moreover, in a fast exponential way.
Today, I am more mitigated about the purpose, and I even dare to think that long term traders (algo or human) can be the big players of tomorrow! And here’s why:
1.In this topic, I will focus on short-term / scalping algorithmic strategies only, since:
- There’s not point / edge against a human to use an algorithmic execution to trigger a 6 month period trade - excepted if you have gigantic sized position to build silently into the market on a specific price and/or in a specific delay - wich is not what we are talking about when it comes to DARWINs, and DARWIN providers.
- There’s a common rule that says “time = uncertainty” (another interesting topic to develop...) and algo traders have gradually pushed this rule at its paroxism.
2. Also, I add that we are only talking here, about trading strategies that take (market) liquidity. Liquidity providers (limit) will have to face different issues.
On a separated note, please, it not a statement against algo developers!
It is only a reflexion about the future.I have a lot of algo traders in my portfolio, and it performs well thanks to them
1. Technological slavery
To take a common language:
- (IF) / my child is born and I immediately put him on a keyboard, without teaching him how to write manually, how to properly hang the pen, desing cursives, etc...
- (THEN) / he will not be able to write the day after I take the computer out.
One aspect of recent evolutions is that we’ve created machines to execute our daily tasks and efforts… yeah that’s fine… until the point that you lose gradually your “know how”.
We put our effort to build machine that build X, instead of putting our effort to build X. So we are not acquiring the “source” knowledge” anymore, but instead we learn how to build an automated process that do it for us to infinite.
The result is simple: dependency!
We can argue that the “know how” is not lost but is transferred from one region of the world to another, depending on very long term (centuries) economic cycles… but…. It’s not the topic and for example, internet is not anymore a choice guys, it has become a prerequisite to manage our lives.
(IF) / for X or Y reason, we take off his computer to an algo trader
(THEN) / game over!
(IF) / for X or Y reason we take off the computer of a manual trader
(THEN) / he can even send his order by postmail and still make money
The trader has acquired knowledge of the market.
The algo developer (as smart as it is) has only acquired the knowledge of designing an hyper-performant tool, to trade the market (= indirect knowledge).
It’s like, instead learning how to fish, you are learning how to build fishing rods.
Can you expect to fill your hunger the day you will be alone, facing the river?
That’s the point again: dependency.
From this note, comes another aspect, directly related:
2. Speed we need!
If speed was an advantage at the early stage - when there was still a lot of room for improvement - today speed has become, not by choice but by competition, an addiction.
At a point that nowadays, only the friction created by successive entries/exits in the market can annihilate a (supposed) robust and profitable strategy if the spread was systematically widenned from 02 to 04, on a 6 months period of time.
Algorithmic trading strategies have gradually shrunk so much their “timeframe field of operation” that they are now competing for crumbs, for which it is not good to take a risk that will be automatically, over time and friction, asymmetrical (my investor portfolio is the first witness )
So the first direct result is: If algos have an advantage over human on speed But this exact component is a crucial information that a discretionary trader can exploit. “If I cannot compete in speed, and, I know that algo operate systematically on short timeframes, I just have to enlarge my timeframe enough, for algo activity to be “inefficient”.
And the second point is: this will lead to the CORE structural weakness of algorithmic scalping strategies - the real topic of this thread:
3. When scalability hurts
"Do you know this bird?"
Yes, exactly… That’s him, Woody Pipspecker!!
My personal vision of algo scalping strategies is that they are like well prepared car, ready for a run, all optimized, extremely performant, brilliant and touching the perfection, but… you have to lighten all the rest of the components at maximum, each piece of textile is weighted and verified. Because each and every gram you add, the more performance you lose.
From what I observe on Darwinex:
The new wave of algorithmic traders and MT4 EAs developers is totally incredible! I’m everyday amazed by what some guys are able to do. I mean, it has nothing to do with the time of Fapturbo it’s a new world!
More and more traders are coming, and good traders if we evaluate roughly the speed of global quality improvement in the DARWINs’ database.
But, more and more INVESTORS are coming too!
This creates the behaviour that we are witnessing actually: fear of missing out!
Once a DARWIN is migrated with a 75 D-Score… you can expect to see it in the “Most Invested” filter in 3 weeks at max.
Because now, Darwinex is a reality. It just… works!
But from this success, comes a reality.
Most of these strategies are algo strategies that cannot aggregate more than 3MM.
Oh yeah, for an independent / retail trader, 3MM with a decent year is far enough to make a living.
But what happens when demand is stronger than supply in a finite liquidity environment?
DARWINs have to pause because no one is making money anymore!
And as a consequence, it produces a sub-issue, as an investor you have to rebalance on another DARWIN (wich will be the next “star” to become sterile/ or paused as others investors are moving too).
Would you rather…?
If we assume that a trading / investment journey is more likely to be a dangerous travel in mountain, than a ballade on a smooth road of bitume. If you want to start from a point A to a point B and this environment,
Would you rather:
Take the optimized sport car and prefer speed over security / certainty
Take the 4x4 car and prefer security with more certainty, over speed
Said differently, in term of hypothetical risk vs expected return, would you prefer to invest 1 full year on a DARWIN (- ie, both same monthly VaR%):
- That delivered 60% return vs.15% drawdown, and -1% monthly divergence in average the last year, or
- That delivers 40% return vs. 20% drawdown, without any divergence
And a more interesting question is, on wich one would you prefer to keep your funds allocated the year after, if you had to keep one?
This is where human discretionary traders will have to play a major role in Darwinex expansion and Evolution.
They can be the BIG players: when/ if investors get tired to jump from a DARWIN to another because they yield sterile returns after 3 months growth in AuM, they will realize that traders/managers still exists, not only developers
Well… It is true that long term DARWINs (mostly human traded) are recognizable because they tend to be less “brilliant”, or sexy. But they are SCALABLE!
And that makes ALL the difference for passive buy & hold, and/or big size investors, and produces an edge over the time against Woody Pipspeckers.
If you can deliver a decent performance year after year, it is not abusive to think that, at equal statistical risk, performance of your investors will be better than those invested on Pipspeckers, if we evaluate carefully all the elements, and project them on a future 6 months to 1 year investment period.
If we take a simple scenario: I invest on 1000x2 traders for a quarter:
- Trader 1 has made 10% performance without divergence
- Trader 2 has made 0% performance with a divergence of -3%
- I’ll pay 20 EUR to trader 1
- I’ll pay 00 EUR to trader 2
- I’ll lose -30 EUR on divergence Trader 2
My net P&L is EUR 2050, so I have paid the equivalent of 50% of my profits at the end.
Don’t forget that Investors’ AuM is just starting to grow on the platform.
Coupling these informations:
- Growth of AuM on the platform
- Lack of scalability on most of the popular strategies
- Most of these strategies are algo strategies
You can reasonably conclude that:
- Algo strategies will have to evolve to manage a better scalability
- Discretionary and long term traders are here to stay!
Each methodology has it pros and cons, and my personal investor view is to mix the behaviors in my portfolio.
Algo strategies find their weakness in their advantage: they traded up scalability against performance - sports cars.
Human traders have not lost their edge. Their edge is just not the same as before: it has moved from performance to scalability - 4x4 cars. Or at least, it is my conclusion on this topic
Algorithmic strategies intimidate you?
C’mon traders, show the fangs!!
There’s no point!! Algorithmic - 45° climbing equity curves - scalping strategies become… just a (yes...very) beautiful illusion after 3MM of AuM!
Looking forward your reflexions on the topic my dear colleagues, algo developers and manual discretionay traders!!
PS: The crucifix is already planted in the ground, just in case