Can't measure success without setting a Goal..or two...and you are gonna need a plan.


sven

Member
So after heading over and posting Introductions: Hi there time to get serious? I made it pretty clear that I'm looking to 'succeed' and frankly been pissing about for years.

WARNING: long long post but hey you don't need to read ;)

Thats all good and well but what is success in trading for me? What the end goal?

Most people will say loads of money etc etc etc and of course that sounds cool and all that but its not really a goal, more of a pipe-dream. We need goals and targets to get anything done over the long haul.

We all have long term goals and short term. Its entirely possible and probably desirable that your goals change over time but that should be planned not made up.

I'm sure we've all got stuck into something, given it 110% for 2 weeks and then the novelty wears off. This is where plans and goals come into it. The idea is to have a mental framework to prevent procrastination and to make incremental progress even when you don't feel like it.... something I've mastered(procrastination and distraction that is)

Get S.M.A.R.T about your goals.
So whats this bull****t about? Well if you are actually going to do something you should have a good ideal about the following. On the off chance you are still reading there are many ways to set goals but SMART is one of the more intuitive ones. You don't need to write it all down but it helps to think of these 5 points when you do.

Specific
What the hell are you actually going to do? 'crack trading' nope.... 'create a top 3 list of trading platforms for my plans' ..yep.

Measurable
Can you tell when you are done? if not then you need to redefine it. Being measurable really helps with motivation. Knowing you are knocking goals off and getting closer

Achievable
Do you have the skills to do it in the time allotted? if not then we need to plan for and allow that to happen first.

Relevant
Clearing goals takes motivation. Motivation is much easier when you are doing something you are interested in and that will directly relate to you.

Time-bound.
Ideally we set a time limit. If you can't make a reasonable estimate on the goal then its possible you don't know enough about how to achieve it or the scope is too much. If you work 9-5(lol I wish) and travel(boo yes) with a family(yey babies..delicious) time is the single most valuable thing you have.

You don't need to be using some tool to do this or a rigid framework but understanding the above when you jot down your post-it notes should really help to make the goals attainable. Well its helps me.


Some High Level Goals.
I basically have 2 long term goals and the main one is;
A portfolio of uncorrelated trading systems across a wide range of Asset classes, instruments and timeframes.
Great but its a bit wishy washy so let me try and SMARTen it up.

Specifically.
A portfolio of profitable trading systems that are robust to shifts in market conditions and sufficiently tested. Timeframes and holding periods will be weeks, days and intraday. Costs and correlation will be the primary criteria for inclusion into the portfolio with a focus on systems that improve diversification. Risk will be managed at all times.

Measurability?
I will measure success by progressing towards that ultimate goal by completing goals along the way. Failure is being distracted from that ultimate goal or deviating from risk parameters.

Achievable?
I do not have the skills to achieve this goal. Thats just a fact. I will need to formulate a plan to identify what skills I'm lacking and attempt to attain those skills and progress. I will ensure progress by planning incremental milestones that move me toward my goal.

Relevant.
I work, have family and I travel. I can't dedicate 24/7 to this but I do have spare time for research(travelling) and software dev skills. This is something I really want to succeed in.

Time-bound.
I don't believe that this journey will ever end but we need to have some limits on time else there is no value in it. So, whilst I'm expecting this to take forever-ever I'm committing my spare time and mental bandwidth for the next 12months.

Any one of the above could be broken down into a million goals and more detail but this is about setting a high level goal.

A year is a long time if focussed. I'm under no illusion that the markets are a lifelong path. I'm not starting from zero but I'm no where near 10.

So 12 months and re-evaluate.


Goal 2, is to improve my discretionary trading skill.
There is a saying;
If you try to chase 2 rabbits you catch none -- a dog.
I believe in eating your own dog food. That is, you should be forced to use the crap you make as you gain insight into its short comings. I believe that you can't really see new opportunities unless you are in the markets looking at charts.

That said, I realise its a different goal and in many ways conflicts with achieving my ultimate goal. So, I'm going to just leave it as that and not commit. Rest assured I'll be still trading but I've decided short term to not try to 'fix' it. I'm going to comeback to it formally at some point in the near future.


Where am I starting and whats the rules of the game?
So we know where I want to be but where am I starting from? Whats the restrictions if any?

I have dev skills. I've 'messed about' with MT4 for a few years early on. I've messed around with jupyter notebooks and python forever. Coding I think I'm good.
Trading skills. I've traded on and off for ever. Never really getting consistent. I understand the basics along with risk and position management and I'm comfortable putting a trade or two on.

The main restriction I have is I can't trade anything listed. No indexes, equities, commodities. Basically I can trade Forex + probably gold. Well thats put a dent on the goal hasn't it.

Yes and no. Next 12 months sure I'm limited but limitations can spark creativity and focus the mind. There are more than enough opportunities in Forex I just need the skills and tools to get me some of them.

Who knows what job I'll be doing past 12 months. Anything other than my current one and those other markets can open up to me. So..may as well crack on and be in a position to add them if the opportunity arises.


Whats the Plan then Sven?
So we know where I want to be and we know where I am... but so much unknown.. so many places to get distracted.

We need a hero..we need Strawman!!! a plan that’s meant to be knocked down because frankly I don't know what I don't know but you've got to start somewhere aka bollocks

I'm not going to put dates but you can assume a couple of weeks for simple things. I'll set a timeframe when before I start the next item.. That will give enough time put stuff off a little for other priorities but still make progress.
  • Platform choice.
    • Define requirements
    • Research the options
    • Evaluate shortlist
    • POC
    • Choose.
  • Broker choice.
    • Supported brokers for platform.
    • Additional Api requirements?
    • Trust
    • Funding.
  • Define System Pipeline
    • High level approach to Algo research, dev through to deployment. Timeframe/instruments etc. How to evaluate, discard, test deploy etc. How to monitor post live!!, how to integrate new systems into the portfolio.
    • Risk parameters and kill switch.
  • Algo!!!!
    • ideas and the pipeline.... loop loop loop.. Go live!!!... monitor as per pipeline.
    • GOTO Algo...

Right anyone that is still alive or remotely interested...

Thoughts? Feedback on the strawman plan? I'm here for ideas. Don't be shy.. I can take it.. Lets get cracking.
 

jack

Administrator
Staff member
Lots to go though here, awesome post! :)

The SMART goals method is excellent and I've recommended people look into it many times. Great place to start when it comes to structuring your plan.

Generally speaking, many people fail at planning out their dreams / goals and this holds them back tremendously. Also, many people don't realize this, but their reluctance to properly plan out the steps needed to achieve something they desire could also be part of their own fear/stalling to accomplish it. They think systems don't work for them, or planning isn't needed, but in reality they don't want to face the potential of failure so by not clearly defining the steps toward their goal it makes the chance of 'failing' non-existent. The human mind is a trip sometimes... lol

Regarding this:
A portfolio of uncorrelated trading systems across a wide range of Asset classes, instruments and timeframes.
I'm not saying it's a bad goal, and heck even I had this in my sights years ago, but in my experience when it comes to trading and edge I found that focusing resources on a specific edge where you might find abnormal returns is more effective than spreading yourself thin working many strategies over many assets.

Example from this site:
https://fxgears.com/index.php?threads/liquidity-providing-and-pure-arbitrage-with-xrp-and-other-crypto-on-coinfield.1215/

When I first discovered the arb in the link above, I started working that trade with my sole attention after my regular work hours for about a month and a bit.. since at the time, it yielded the greatest risk adjusted returns for my time over any other action. Burning the candle at both ends catches up to you after a while and I found other more profitable uses of my time during my non-day job hours, so I reduced how much attention I was giving that strategy, but for a while it was my "evening entertainment" without much exception.

Another example is a buddy of mine who worked a SPX / VIX options arb trade on the CBOE for years that took advantage of a structural advantage in that market he had over the institutions making markets on the underlying products... buddy couldn't care less about directional trades or patterns, only worked his book, edge, and ignored all other aspects of the markets. He made quite a bit of money off that trade and only stopped once the CBOE changed a few of their trading rules that blunted his edge hard. He now exploits another edge in another market and is still kicking ass. Single ideas, focused on, and leveraged, made this guy's career. He never has to worry about money again.

Anyway.. not saying you're going about it wrong. I'm just saying don't be afraid to focus on one type of system if you detect an edge that's head and shoulders above what would be considered 'decent' returns.


Finally:
Whats the Plan then Sven?
Consider your order is reversed. Brokers are just tools. You find the best tool for the job. You have to define the job first, then find the right tool.

Define what kind of edge you might be able to find given a specific type of data set. Like, if you find daily bar data with volume, and you want to test against it, you're looking at: bar patterns, volume or time analysis (volume profile, TPO, etc..), indicator based ideas derived from price bars, general time series analysis, seasonality, etc...
However, what you can't test for is stuff related to the depth of market, frequency of trades, rates of change in price over time, derivative/options market impact on the underlying, etc...

So acquire a data set, come up with an hypothesis based on your observations of how the underlying product trades which can be verified by the data you acquired, write code to verify said hypothesis, if proven then find a broker which trades said product and has an API you can connect with (if the API is even required, as trading on some timeframes might be just as easy to manually manage the system,) then start trading.

I mean, it's easy to waste a lot of time in the power tool aisle of your local home improvement store while daydreaming of all the ways you could use a new band saw... but unless you've got wood and a blueprint waiting for you at home you're just wasting time.

Hope that helps.
 

sven

Member
Thanks for the comprehensive response Jack and for your insights

reality they don't want to face the potential of failure so by not clearly defining the steps toward their goal it makes the chance of 'failing' non-existent.
This is something that would sometime plague my creation of ideas. I would glance through a chart and almost be scared to backtest it...incase its garbage and I fail.

On the portfolio stuff..
I'm not saying it's a bad goal, and heck even I had this in my sights years ago, but in my experience when it comes to trading and edge I found that focusing resources on a specific edge where you might find abnormal returns is more effective than spreading yourself thin working many strategies over many assets.

Example from this site:
https://fxgears.com/index.php?threads/liquidity-providing-and-pure-arbitrage-with-xrp-and-other-crypto-on-coinfield.1215/
I read your arb post and it was fascinating. I hear you on portfolio. I guess my point is more one of just working toward more than one edge and being aware of the relationships between them to manage the risk. At the end of the day a portfolio is a pipe dream if you can't exploit a single edge.
Taking your point that searching for an abnormal return rather than spreading thin makes sense for sure.

Anyway.. not saying you're going about it wrong. I'm just saying don't be afraid to focus on one type of system if you detect an edge that's head and shoulders above what would be considered 'decent' returns
:) didn't think you were suggesting It was wrong. At the end of the day I'm here for as many different perspectives as possible. Yours is clear and I can't disagree as I've not investigated. rest assured I will and then I'll be able to decide if its something that aligns with my skills/inclination. Also fwiw.. I'm not really interested in 'decent' lol.. i'll settle for that as a baseline.

On the whole got it in reverse theme...edge then tools.
Define what kind of edge you might be able to find given a specific type of data set. Like, if you find daily bar data with volume, and you want to test against it, you're looking at: bar patterns, volume or time analysis (volume profile, TPO, etc..), indicator based ideas derived from price bars, general time series analysis, seasonality, etc...
However, what you can't test for is stuff related to the depth of market, frequency of trades, rates of change in price over time, derivative/options market impact on the underlying, etc...
interesting point about defining edge, getting data and testing...then finding a broker.

My problem with that..and bear with me. I'm not sure I want to build my own backtest framework. or maybe you weren't suggesting that?
Its such an easy way to disappear down a rabbit hole. So when I was planning on looking at tools MT4/5/Ctrader/backtrader I was really coming from the point of view of 'what am i trying to build (portfolio)' and what is best... ie. if I want multi currency testing MT4 and ctrader are probably out. MT5 and backtrader support mutti instrume testing.

But... i take your point. I'm a strong python dev so maybe I just stick with jupyter notebooks and python with a bit of backtrader but I'm trying avoid NIH syndrome(not invented here ;)

I have seen your stack thread.. its clearly bespoke... is your backtesting also bespoke?

thanks man, much appreciated. Your comments are not wasted.
 

Top