Wyckoff trading method, page-643

@Jako64

I understand where your coming from - I enjoy charts myself.

For interests sake, the reason I focus on code is, whilst I like charts, what I like more is time on the water and time away from the screen.
For me, the best way to get out there is to sit in front of a screen and teach the machine to read the charts (A bit ironic).
I also believe machines can become better traders then humans.

My ethos is;
Trading decisions should be binary & rule based.
If a human can visualize it, it can be programed & the best way to ensure rule based trading strategies are adhered to is teach a machine to do it (Machines don't make mistakes or get creative unless they are programmed that way).
[Perhaps I could get a job at Cyberdyne ]

There are a lot of ASX charts & to analyse them all manually wouldn't leave too much time for sleep.
If I were to only analyse a number of them I would miss many high probability low risk: reward setups elsewhere in the market.

For mine a major part of trading is estimating risk: reward before entering a trade (nearly all retail traders I meet do not have a methodology to calculate risk or base their risk reward or rely on a single number - the back test results of their TA system)

When a machine makes a buy/sell decision, it just needs to record a static copy of the variables (metadata) that brought it to making that decision & a static copy of the estimated risk: reward.

A dashboarding layer showing historical win:loss & profitability broken down by trading plan provides peace of mind that the trading system is solid and proven.
If the machine is managing the portfolio, with the right code it can also become adaptive & self-learning to add weight to the setups that have the best track records.

My trading engine analyses every active ASX chart (1828 of them as at today) every 30min
It provides a list of buy recommendations ranked highest to lowest based on calculated risk-reward.
If I buy, I tick a box that tells the machine I bought (now it will provide sell recommendations on that trade).

Whether I participate in a trade or not, the machine records all proposed trades (both buy & sell) in a trading diary (populated with the trade details such as buy/sell/reason and detailed metadata as a snapshot of the variables that went into that decision).


As the model tracks multiple trading plans the reporting layer also tells me where attention is required to improve or cut a particular plan. The result being a self-analysing, fully automated system without error brought about by human input (except the programmer ).


8 years back in the name of free time, I took my trading hat off and put my investor hat on.
Now I'm in the process of migrating my old quant & some new into a new model on a new SQL platform (hence the posting activity on HC). I've still got a few months coding before it's BAU.

To your comment about flexibly interpreting data, I notice you mention excel & VBA in a prior post; VBA is one of my languages & I've programmed a number of adaptive and self learning models. If you have a specific question feel free to post it on my quant thread. If I can answer without spending too much time I will & if I like where your going I may just write & hand you the code and keep a copy for myself