Sim Against Live
A simulator teaches mechanics well and cannot teach either cost or consequence. It sets to zero every number this module spent six lessons measuring, and it removes the one thing that makes a loss a loss. Knowing exactly which of the three you are practising is what makes the tool worth using.
Prerequisites: Lesson 10, for the four charges the worked example puts back, and lesson 12, for the instrument it puts them back on.
What a simulator is genuinely good at
Start with what it does well, because the case against it is usually overstated and the overstatement is what makes people ignore the case. A simulator will teach you your platform: which order type does what, where the stop field is, what happens when you fat-finger a quantity. It will tell you whether a setup exists often enough to build a week around. And it will tell you, honestly and cheaply, whether you can follow your own checklist twenty times in a row.
Those are real skills and you keep every one of them. They are also, all three, questions about you and the software rather than about the market, which is the clue to where the tool stops.
The first gap: it fills you for free
A simulated fill is a fill at the price on the screen, in full, immediately. Every one of those three words is a gift the market does not make. Lesson 4 said the spread is the price of immediacy and lesson 11 said your fill depends on the size resting in front of you; a simulator charges you for neither. It takes no commission unless you tell it to, and most readers never do. It charges no financing on an overnight position.
There is a subtler version of the same gift, and it catches limit orders rather than market ones. A simulator fills your limit when the price touches it. Live, a touch is not a fill: lesson 3’s queue is standing in front of you, and price can trade at your limit repeatedly without ever reaching your place in the line. So the tool is generous twice over — it fills your market orders at a price nobody offered, and it fills your limit orders on trades that never got as far as you. The second is the one that quietly inflates any strategy built on resting orders, because in a simulator such a strategy never misses.
So the simulator does not merely flatter your results. It flatters them by exactly the amount this module has spent six lessons teaching you to measure, which means you already have the number needed to undo the flattery, and the worked example below is nothing more than putting it back.
The second gap: nothing is at stake
The other half is harder to quantify and easier to state. In a simulator a loss is a number that changes on a screen. Live, it is money that was yours and now is not, and the reason that matters is not motivational: a loss and a gain of the same size do not weigh the same, and the asymmetry is one of the most reliably measured findings in the study of how people decide under risk. A simulator sets that asymmetry to zero along with the commissions.
What follows is narrow and worth stating carefully. The simulator is not failing to teach you discipline; discipline is not the variable. It is failing to present the input that discipline is a response to. You cannot rehearse how you behave when something is at stake in an environment where nothing is, and no amount of additional simulated trades changes that, because the missing quantity is not in the trades.
What does teach it
Which points at the remedy, and the remedy is not more simulation. What the tool withholds is a stake, so the way to get one is to take one — the smallest your broker will accept. The real instrument, real money, at a size where losing all of it is an amount you would not think about twice. The mechanics carry over unchanged. The costs stop being assumed and start being measurable, which is the arithmetic below done for real. And the consequence is present at a scale that teaches instead of damaging.
The argument for the middle step is that fear at a few dollars a trade teaches exactly what fear at several hundred teaches. The difference is only that one of them leaves you an account to apply the lesson to. What not to do is step from a simulator straight to full size, because then your first experience of real money and your first experience of real size arrive on the same trade, and afterwards you will not be able to say which of the two you were reacting to.
Which makes it an instrument test
Put those together and the tool has a precise job. A simulator answers “can I operate this?”, and it answers it well. It does not answer “does this make money?”, and no run length fixes that, because the shortfall is structural rather than statistical. Treat a simulated equity curve as evidence about your hands and never as evidence about your edge.
Putting the costs back
Take the mid-cap from lesson 12: $40 a share, quoted five cents wide, moving about 250 basis points on an average day. You buy 250 shares, a $10,000 position, and you use a 1 per cent stop, so $100 is at risk on each trade. The simulator reports a hundred trades at a 55 per cent win rate with wins and losses the same size, and it puts you up $1,000. That is the number people take to a live account.
Now put back the three charges the simulator dropped, using this module’s own figures. The round trip crosses one full spread, which on 250 shares at five cents is $12.50. Commission at a flat dollar a side is $2.00. Slippage of a cent a share beyond the touch, which lesson 11 would call modest for this size, is $2.50. Call it $17.00 a round trip, against $100 of risk, so lesson 10’s ratio is s = 17 ÷ 100 = 0.17.
| Over one hundred trades | What the simulator reported | What the same hundred trades net |
|---|---|---|
| Win rate | 55% | 55%, unchanged |
| Gross result | +$1,000 | +$1,000, unchanged |
| Cost of trading | $0 | −$1,700 |
| Net result | +$1,000 | −$700 |
| Win rate needed to break even | 50.0% | 58.5% |
Read the last row first, because it is the one that decides. With no costs the breakeven win rate is 50 per cent, and 55 per cent clears it comfortably. With the costs back it is 58.5 per cent, and 55 per cent does not clear it at all. The strategy did not stop working somewhere between the simulator and the broker. It never worked; the simulator was simply not charging for it.
And notice how little of the loss is dramatic. There is no gap, no revenge trade, no bad week — nothing but seventeen dollars, a hundred times, on an instrument that lesson 12 ranked in the middle of its table rather than at the bottom. The most expensive thing in this lesson is arithmetic that a spreadsheet does in a second and a simulator declines to do at all.
What this does not settle
That simulators are useless, or that you should skip one. The opposite follows: because it answers a narrow question well and cheaply, it is worth using for exactly that question. What does not follow is treating its equity curve as a result, and the correction is not to simulate for longer.
Nor are the specific costs yours. The $17.00 is built from one instrument, one size and one commission schedule, and every part of it is a number you were shown how to measure rather than one to import. Run the same $10,000 through the top of lesson 12’s table and the breakeven moves from 50 per cent only to 51.2 per cent; run it through the bottom and it moves to 318 per cent, which no win rate can reach. The method transfers, the seventeen dollars does not.
And the second gap is genuinely not quantified here. This lesson prices the cost side exactly, because the cost side is arithmetic, and it declines to put a figure on the other one. Anyone who offers you a precise number for what fear costs is estimating, and a lesson that has just spent a page on the difference between a measured cost and an assumed one should not end by assuming one.
And the three columns are not as separable as sorting them into three columns suggests. The first one was described as fully bankable: mechanics carry over. They carry over as motor habits, and following a checklist twenty times with nothing at stake is not the same act as following it once with money on it, which is the whole point of the second gap turned back on the first. So even column one comes with a discount this page cannot size, and the only honest thing to say about its width is that the smallest live position you can open measures it and nothing else does.
A simulated track record is evidence that you can work the platform. It is not evidence about anything else, and the arithmetic that separates the two takes a minute.
Problems
- Re-price your own simulated record. Take whatever result your simulator has given you and count the round trips. Multiply by your own cost per round trip, built the way this lesson builds it: one full spread, your actual commission schedule, and a slippage figure from lesson 11 rather than from optimism. Subtract. If the sign changes, you have learned the whole lesson for the price of an evening.
- Find the win rate you actually need. With wins and losses the same size, the breakeven win rate is one half of one plus s. Compute yours, then compare it against the win rate your simulator reported. The gap between those two numbers, not the equity curve, is the thing to be pleased or worried about.
- Sort your own list. Write down every claim your simulator has made about your trading, then put each into one of three columns: mechanics, cost, or consequence. Everything in the first column you may keep. Everything in the second you can repair with arithmetic you already have. Everything in the third is still ahead of you, and the point of the exercise is to know which of your confidence belongs to which column before an account finds out for you.
Sources. Brad Barber and Terrance Odean, “Trading Is Hazardous to Your Wealth” (Journal of Finance 55, 2000), which separates the gross returns of tens of thousands of retail accounts from their net returns and locates the underperformance in the gap between them. Daniel Kahneman and Amos Tversky, “Prospect Theory” (Econometrica 47, 1979), for the asymmetry between a loss and a gain of equal size, which is the quantity a simulator removes. And David Bailey, Jonathan Borwein, Marcos López de Prado and Qiji Jim Zhu, “Pseudo-Mathematics and Financial Charlatanism” (Notices of the AMS 61, 2014), on why a simulated record that was selected for its result is weaker evidence than it appears.
That closes the cost of trading. You can now price a round trip, rank an instrument, name the object, compute the level at which someone else acts, and undo a simulator’s flattery. The next module asks the question all of it was preparation for: what an edge actually is, and how you would know you had one.
Every Trade Starts Negative
The four charges a simulator sets to zero, and the ratio they go into.
Read Lesson →Slippage and Impact at Retail Size
Where the slippage figure in the worked example comes from.
Read Lesson →What Should You Actually Trade
The instrument this lesson re-prices, and the table that decides how much it matters.
Read Lesson →When Your Broker Acts Without You
The other thing a simulator never does: hand your position to someone else.
Read Lesson →Educational only. Trading involves substantial risk of loss. Not financial advice. Past performance does not guarantee future results.
💬 Discussion (0 comments)
Loading comments...
Ready to Trade with Signal Pilot?
Apply your trading education with professional indicators and real-time market analysis tools.
Back to Signal Pilot →