CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. -- % of retail investor accounts lose money when trading CFDs with this provider. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.

[Opinion] Discretionary vs. Quantitative Trading

This is an excellent short article, I highly recommend everyone interested in this area reads it ~

And as always, it would be great if Darwinex users shared their opinions on this here on this thread :slight_smile:

4 Likes

100% agree with the autor.
I would substitute “discretional” with “fundamental”.
Saying quantitative trading doesn’t work is like saying technical analisys doesn’t work.

My opinion is that fundamental trading works only with stocks.
With currencies and commodities it is better to stick with neutral and technical trading.
Technical trading can be managed in a manual or automated way.

3 Likes

@CavaliereVerde I personally would not substitute discretionary with fundamental. Discretionary in this context just means that a human collects and analyses information on which he bases his trading decisions on. This information could both be technical or fundamental.

In my opinion, probably the strongest approach is combining both. As machines and algorithms are clearly superior in analysing and clustering huge amounts of data whereas human experts can oftentimes better evaluate and assess this information.

However, saying that quantitative trading cannot be successful whatsoever is totally beyond me. There is a multitude of examples that clearly show otherwise. So yes to “defense of the quantitative approach”

4 Likes

Thank you @integracore2 for having spotted this Article, read it all just now. :grin:

Totally agree with you @CavaliereVerde . :thumbsup:

Quantitative traders takes a trading technique from ideas developed on charts and subsequently they create a model to automatize it according to its operativity into historical market data. No way to say that a simple technique ( IMO the best ) who often works and statistically has more profit than loss can’t do money into future if good optimized and not over-fitted in past. Another piece in favor of ALGOs is that they delete any kind of emotion from charts, and no matter how trivial is this, but I think that is the base of many losses accumulated from the rest of traders.

Discretionary-trading is a decision based type of trading, systems who work according to the decisions of a person are susceptible from the psychology of the trader and from my experience it ends often to not follow the initial Trading Plan. :grin:

VolcanoFX ( GTD )

3 Likes

totally agree with everything that has been said so far.
Big Yes to quantitative/algo trading. Not only algo trading allow you to create and test many strategies, it also allows you to monitor a lot more currency pairs in different market regimes and as vanilla or as exotic as you want, 24/7. This is an amazing power. The algos can be totally automated, can be used to make suggestions if a trader does not feel fully comfortable with a 100% automated system and the entire R&D, testing, deployment and performance continuous analysis, if done correctly, mimic the scientific method of approaching a highly complex system which is the financial market.
One additional comment is that emotions are very much attenuated but not necessarily eliminated altogether from the equation. Some traders have decided to manually stop their automated system although it was still within their risk management criteria, just to discover that it performed quite good have they obeyed their initial risk parameters. These traders “panicked” due to old-fashion negative emotions, so I think there is a “quantum leap” in the impact of emotions as compared to discretionary trading but, sadly, while humans develop and control the systems, they are still there (just a lot less…)

3 Likes

Curious as to how you decide with an automated system when it is no longer working and should not be used. Do you have a pre-defined draw down limit, or other checks and balances to monitor what is happening. This is a general question, not specifically aimed at @KlondikeFX, but I wasn’t sure where else to post it. I’m not asking for technical information, just want to get a general idea as an investor how these things are prevented fro running wild.

1 Like

Many algo traders have a limit of twice the historical drawdown.
The general rule is that you stop a strategy when the live behaviour becomes inconsistent with the backtest.
IMO simple stategies should not need to be permanently removed, maybe suspended and recalibrated.
The majority of algo traders run a portfolio of simple strategies, so suspending one strategy doesn’t have a heavy impact to the performance of the portfolio.

1 Like

That’s actually pretty difficult. I don’t think there is a difference between automated and manual systems in that regard. In my view this is a completely discretionary decision which comes down to the trader’s knowledge and experience. I don’t have any automated checks or monitoring systems, nor do I have a fixed threshold in terms of drawdown or stagnation that lets me decide that the strategy is no longer working. I do, however, analyse most trades afterwards and I do check the general behaviour of the underlying currency pair.

E.g. I stopped trading GBPUSD on most of my accounts after the referendum because I was not sure if the future behaviour would still be in line with the backtests.
I am still running my algos on a couple of smaller accounts to reevaluate this decision afterwards, though. After Q1 2017 I will do a deeper analysis of the trades during those 9 months and decide whether I will resume trading, prolong the pause or abandon my current algos for that pair.

6 Likes

I think at this point, seeing how HFD is “deep-diving” into another 10% loss month, Darwinex needs to realize that there’s not much predictive value in neither D-score nor Darwin Quote Data.

I also believe that Darwinex is missing the “big picture”. What I mean is:

  1. Darwinex is a data driven company. Do you have any data on Darwins (or Underlying Strategies) Returns of Manual and Semi Automatic vs Fully Automatic ones? I think the investors would greatly benefit from this if data proves that either one or the other has a higher chance of bringing returns. At this point, there is no indication of this when making an investment decision in one of the Darwins.

  2. No extremely good MANUAL trader will ever trade for Darwinex or any other brokerage firm, they will only trade for themselves because if they are so good, they don’t need any 10-20k a month from investors. The very good traders can most likely do that money themselves without having to risk someone else’s money - notice how I am saying TRADER and not investor!

  3. Neither manual nor semi automatic strategies should be welcome on Darwinex. If Darwinex should be focused on anything, it should be exposing the best fully automatic strategies, such as the ones that TradeSignalMachine is producing. Other than THA, how many Manual traders with decent Returns are here on Darwinex? - This “statement” is assuming that what I asked in point 1 proves that Fully Automatic strategies with x amount of track record outperforms manual or semi auto ones with same amount of track record.

3 Likes

Sorry, I strongly disagre! I can‘t even read what you just posted it is so idealistic! Every automated strategy is only as good as its creator who wrote it. You on the other hand, are cherry picking one Darwin (HFD) example to prove your point. Its the same idiotic view ETF Investors have vs stock picking. In my eyes, the only thing you have proved is disqualifying yourself!

2 Likes

I’m sorry, but what you say is not coherent, and it seems more like a joke to generate a discussion that makes no sense, I hope that apart from me, no one else lives up to your comment.

2 Likes

HFD had at some point 4 mil AUM. Why don’t you ask those investors that had a huge DD if it’s an idiotic view or not? Of course I am cherry picking but I am not cherry picking one darwin with no AUM, I am cherry picking the one with the biggest AUM at some point in time. You think those investors are unhappy just with HFD or ALSO with Darwinex?

By calling my view idiotic without giving any arguments and without waiting to see any data, like the one I asked for, you are disqualifying yourself. You sound like those people that believe in something so strongly that every other belief MUST be stupid, or else they feel bad about themselves.

My man, how about instead of asking how good is Darwinex to the investors that made money and the trades that made money, ask about Darwinex to the ones that Lost a ton of money to HFD and others?

I can see now you got some skin in the game in https://www.darwinex.com/darwin/IYU.4.10/ which gives more sense as to why you got so upset without providing any arguments to prove your point.

You’re a funny guy, with a great view, and every other view is idiotic.

1 Like

A joke to generate a discussion?

My man, HFD had at some point 4 mil AUM. How about you ask those investors that incurred a 20% loss if it seems like a joke to them?

You, like your friend above, Sterso, give no arguments.

Why in the world would I be upset?:joy:
You claimed something baseless, that automated strategys are „the only way to trade“ because of one Darwin to prove your Point! I only contradicted you, by emphasising that automation is not necessarily better, due to it only being as good as the creator who wrote the code. That is a proven fact!

I‘m not interested in idealistic views (like yours pro automation) neither am I a missionary trying to force my view on others! So please take a deep breath, think first then write. Of course I have skin in the game, why wouldn‘t I as a Trader…

You got quite upset, called my “view” idiotic and provided no arguments, but it’s ok, it means you’re competitive.

Again, it’s not an idealistic view, it is a REQUEST to compare the strategies at Darwinex. To compare the Manual and semi-auto to the Fully Automatic strategies.

So please, take a deep breath, read more carefully what I wrote and then write.

1 Like

The disctinction between manual semi-auto or full auto is not so granted.
There are traders that generate a trackrecord, DScore rates that trackrecord.
The efficiency or predictive value of DScore has nothing to do with these distictions.

1 Like
  1. My point is that there is No indication (granted by Darvinex) to the investors that are making an investment decision of a strategy being manual, semi or fully algo. They’re a startup, they have to find a way to make it “granted” and to give some indication as to Either Manual, Semi or Fully Algo generally bring “better” results.

  2. Predictive value of DScore has nothing to do with manual, semi or fully algo strategies. I was merely pointing out how pretty much useless the DScore is at this point. Nothing in that DScore or in the investable attributes could have predicted 2 months of 10% loss in the past 6 months.

So you seem to believe that algo is better than manual.
I am an algo trader but I cannot sustain that.
Besides THA we have ULI and UYZ , ULI is managed by @javicolonbo that is a founder and a quant…

That was my same point! I‘ve been trading for about 10 years now. Never have I come across an automated Strategy that works all the time long term. Such things just don‘t exist. Its a mirage, and always down to the Person behind the code/strat.

In my Personal view, due to andrei calling me out, it dosn‘t really matter how a trader does his execution. Trading was, has and will always be atleast 80% mental strength and to a lesser extent strategic. Of course this all takes into Account that said trader has specific Risk & Money Management rules in Place! Otherwise a potential blowup is inevitable (as Thanos would put it)

But as i said above, that ist just my Personal subjective view on this Topic. Never have i claimed to know it all, neither would I chose one above the other. All types of execution have their justification for existence.

PS. please excuse my spelling, German is my first language

1 Like

What I believe is that there could be a point where a Manual trader is so good that he does not need investors money. The same goes for the Algo traders. There could be a point where the Algo trader is so good that he does not need investors’ money.

At this point in time, we don’t know whether Manual (or semi manual) traders on Darwinex are outperforming Fully Automated strategies and I find it really weird that Darwinex did not yet try to gather any data on that.

Yes, you are right, besides THA there is ULI and UYZ and probably others. Let us then take THA, ULI, UYZ and many other manuals and compare them to Fully Algo.

It’s improbable that there’s no valuable data there and if by some chance there is no data, I still strongly believe that Investors should be aware when they make an investment decision that This Darwin is Fully Automated and This Darwin is Manual.