Should You Trade a System by Hand or Automate It?
· #systematic-trading #trading-psychology #trading-automation #expert-advisor #gold-trading
It depends on which one is your bottleneck. If the hours when your setups form collide with your job or your sleep, no amount of discipline fixes that, and automation is the honest answer. If you can be at the chart but you tend to override your own rules, automating only moves that fight somewhere else. Then it is better to keep the click and put better structure in front of it.
MasterVP · Systematic Trading Journey
Written by the founder of KenKem, a software engineer of twenty years who traded gold by hand before building the research stack behind MasterVP. As a self taught quant trader I built the same method twice, once to be read on a chart and once to be executed, and this is how I think about which one fits whom. This is knowledge sharing, not a recommendation to trade anything. Every figure below comes from my release 1.26 write up of 15 September 2026, measured on real tick data, and none of it is a live result.
Why is this a calendar question before it is a skill question?
Because a setup does not wait for you to finish your meeting.
On my own gold systems the trades are short. In the release 1.26 backtest the median hold was 23 minutes on the five minute system and between 2.2 and 3.3 hours on the ten minute one. A trade that lives for 23 minutes has to be seen, judged and entered inside a window that is easy to miss if you have a job, a family or a time zone that puts the busy session in the middle of your night.
That is not a character flaw. It is scheduling. An indicator in that situation mostly shows you what you missed, and a trader who keeps seeing what they missed tends to start chasing the next one. For someone whose calendar collides with the market, automation is not the advanced choice. It is the choice that matches their life.
Why do the trades you miss matter more than they look?
Because in a system like mine, the few big trades carry everything, and the trade a human skips is never random.
My results are concentrated. A small number of large winners carries the whole record, and one trade that reaches its full target pays for about five full losses. That shape is normal for breakout trading, and it has an uncomfortable consequence. Five losses in a row is not a sign that something broke. It is ordinary arithmetic.
When I traded by hand, the trade I skipped was almost never a random one. It came after a losing run, when my confidence was lowest and the next setup looked exactly like the last five that had failed. In a distribution where one winner pays for five losers, skipping selectively after losing streaks cuts off precisely the part that pays. A machine does not get tired of being wrong. I did.
Why does automating not fix a trader who overrides?
Because the override does not disappear. It changes shape.
If you have ever closed a trade early while knowing you were breaking your own rule, an Expert Advisor may not cure that. The impulse that used to move a stop can just as easily switch the EA off after the fourth loss, or loosen a risk setting in the middle of a drawdown. I wrote about where that feeling goes in does automating your trading actually remove emotion, and about the override itself in why traders override their own trading systems.
Some people genuinely think better with a chart in front of them, and handing the decision away makes them worse, not better. For them the honest move is to keep the click and make it with clearer structure.
What does the same method in two shapes actually mean?
One research idea, validated once, shipped at two levels of automation.
Behind both sits the Dquants engine I developed, in three layers. TradingView and Pine are where I sketch an idea and look at it. A Python research layer over tick data handles the statistics, the gates and the registry of what passed. A deterministic C++ core holds the strategy logic with no broker code anywhere near it, so the part that decides can be tested on its own.
From there the method ships twice. The Master Volume Profiler indicator on TradingView and MetaTrader 5 draws the value area, the trigger and the filters that passed or blocked, then leaves the order with you. The Expert Advisor on MetaTrader 5 executes the rules and the risk settings you configured.
How closely does a hand-traded version match the backtest?
Less closely than people assume, and I can put a number on how much even two machines drift.
Before anything ships, my research engine and MetaTrader replay the same tick history and have to agree trade by trade. Across five deployed legs and more than 4,900 trades, they agreed on trade direction every single time, and on the exact entry moment between 96.8 and 99.8 percent of the time. Two programs, same rule, same ticks, and they still disagree on a few entries in a hundred.
A person reading the same rule by eye, from a phone, a few seconds late, will drift further than that. I also test execution delay at 0, 5 and 15 milliseconds, because a breakout entry is worst exactly when the move is fastest. I have never measured my own click latency and I will not invent a figure for it, but it is not measured in milliseconds. So I treat a hand-traded version as a close cousin of the backtest, not as the backtest itself. The TradingView version draws the same logic, but it is not a validation environment, since it has no broker tick fills.
What does the Expert Advisor still leave to you?
Everything that decides how much you can lose.
The EA executes on the configuration you choose: risk per trade, the daily loss limit, the overall drawdown limit, the sessions it may trade. Those stay your decisions, and so does the decision to keep running it through a bad month. An alerts layer reports every entry, exit and protective action, so you can watch it, question it and take manual control. Automation takes you out of the moment of entry. It does not take you out of responsibility for the account.
Who should use neither?
Anyone looking for a switch to flip and forget.
The drawdowns are real, the losing streaks are real, and both versions ask you to keep thinking. If you want a tool that promises to win most days, this is the wrong one, and I would rather say so than sell it to you. The results also arrive in bursts, in the months when gold is moving, with quiet stretches in between.
For a lot of people the free editions are enough. They draw the structure, the master profile lines and the value area, which is the part of the method that does the most work. Someone who uses them for six months, never pays me anything, and stops taking impulsive trades is a better outcome than a quick sale to someone who was not ready.
Why is the process the part that lasts, whichever shape you pick?
Because every edge eventually meets a market it dislikes, and only the process notices.
What I am proudest of is not either product. It is the machine that finds, tests, rejects and monitors ideas without fooling itself. As of release 1.26, 158 experiments had been declared in advance with their decision rule written before the result. 140 of them never became a product, and 3 shipped. Hand or machine, the rules you execute are only as good as the process that was allowed to reject them.
Frequently asked questions
Is automated trading better than manual trading? Neither is better in general. Automation fits a trader whose schedule clashes with the market, and manual trading fits one who decides well at the chart. The deciding question is whether your bottleneck is presence or execution.
Can a manual trader reproduce a backtest? Only approximately. Even my research engine and MetaTrader, running the same rule on the same ticks, agreed on the exact entry moment between 96.8 and 99.8 percent of the time across more than 4,900 trades. A human adds reaction delay and selective skipping on top of that.
Why does missing a few trades matter so much? Because many systematic strategies earn most of their result from a few large winners. In my own gold systems, one trade that reaches its full target pays for about five full losses, so skipping the wrong trade can erase a long run of small gains.
Does an Expert Advisor remove emotion from trading? It removes emotion from the fill, not from the person. The temptation moves to switching the EA off or changing its settings during a drawdown, which is why those rules need to be written down in advance too.
What does a trader still control when using an Expert Advisor? The risk. Risk per trade, daily and overall loss limits, trading sessions, and whether the EA keeps running at all remain the user's decisions. The EA executes the configuration it was given.
How long do trades last in a short-horizon gold system? In my release 1.26 backtest, the median hold was 23 minutes on the five minute system and 2.2 to 3.3 hours on the ten minute system. That is short enough that being away from the screen for one meeting can mean missing a trade entirely.
Is a TradingView indicator enough to trade a systematic method? It can be, if you are present when setups form and you follow the rules consistently. It shows the read and leaves the order with you, and it is not a backtest engine, so its historical markers are a study aid rather than proof.
If you want to see where this ended up, MasterVP is my own intraday method built on public volume profile theory. It ships as the Master Volume Profiler indicator on TradingView and MetaTrader 5, for the ones who want to trade manually, and as Master Volume Sniper on MQL5 market, an Expert Advisor that executes the rules and risk settings you configure.
This article was composed from the two-doors posts of the MasterVP build log series. Educational purpose only. Not financial advice. Figures cited are backtest results on real tick data, and past performance does not guarantee future results. Every trader remains responsible for their own risk settings and decisions.