Skip to main content
MARGIN CALLv0.51.0
Guide

What is a stock market simulator?

A simulator generates its own market. That is the whole difference, and it decides what the thing can and cannot teach you.

Last reviewed 20 August 2026

The short answer

A stock market simulator is a program that generates a synthetic market — its own companies, its own prices, its own economy — and lets you trade in it with money that does not exist. Nothing you do reaches a real exchange, and no real security changes hands.

That distinguishes it from the two things it is most often confused with. A paper-trading account sits on top of real, live market data and simply declines to send your orders to an exchange. A brokerage demo account is the same thing wearing the broker's interface. A simulator generates the market itself.

Why generate the market

Because real data only ever gives you one history, and you have already lived through it. If you paper-trade the last six months, you are trading a period whose ending you know, at a speed the calendar sets. You get one recession per decade and you get it in real time.

A generated market can be run at any speed, restarted with different conditions, and made to produce a crisis on demand. In Margin Call a run covers about five years of market time, which means you can sit through a full rate-hiking cycle in an evening and then do it again having learned something.

The cost is that the market is only as good as the model behind it. This is the question worth asking of any simulator, and most of them fail it.

What separates a good model from a bad one

The weak version of a market simulator moves each price by a random amount every tick, applies a drift, and calls it a market. It looks convincing on a chart and teaches nothing, because there is nothing underneath the price to reason about. No amount of analysis can beat a random number generator, so the skill you build is superstition.

A model worth practising against has prices that are consequences. In Margin Call each company has a fair value derived from its earnings and the prevailing interest rate, and the traded price is that fair value multiplied by a mispricing term that decays back toward zero over roughly sixty days. So a stock can be wrong, and being wrong is temporary, and you can form a view about which it is.

  • Prices should follow fundamentals over time, not wander from them forever.
  • Interest rates should reach earnings, not just sit on a dashboard.
  • Large orders should cost more than small ones, for a stated reason.
  • Volatility should cluster — calm periods and violent ones, not uniform noise.
  • Correlations should rise in a crisis, because that is when they do.

What a simulator cannot teach you

It cannot teach you what it feels like to lose money you needed. Every honest account of trading says that the emotional register of real risk is the hard part, and no simulation reproduces it. Position sizing that is easy in a browser game is not easy with rent attached.

It also cannot teach you the specific texture of any real market: which venues fill what, how a particular broker routes orders, how a specific company reports. Those are facts about the world, and a generated market does not contain them.

What it can teach is structural: what an option chain is telling you, why a leveraged company falls further when rates rise, what happens to a crowded position when everyone's stop is at the same level. Those transfer, because they are properties of how markets work rather than of any one market.

Where Margin Call sits

Margin Call is free, runs in a browser, and needs no account. It simulates 108 companies across 12 sectors, plus options, crypto, bonds, foreign exchange and prediction markets, on top of an economy with its own inflation, unemployment, interest rates and business cycles.

It is not a brokerage, it is not connected to any real market, and nothing in it is financial advice. It is a simulation you can be wrong in for free.

In the game

More guides