Skip to Content
Trading in practiceBeginner mistakes

Beginner mistakes

Most of a beginner’s losses come not from bad ideas but from how the ideas are handled: a position without a stop, averaging down, a trade to “win it back”. This is a basics-level lesson: ten typical mistakes, why they are costly and what to do instead. Two of them — changing the rules after a losing streak and overfitting — are shown in interactive examples.

1.Trading without a stop

Without a stop, the loss is limited by nothing but your resolve to close the position — and that resolve weakens with every point against you. A stop loss is placed before the entry, and the position size is calculated from the distance to it (lesson Risk and position size).

2.Averaging down

You bought at 100, the price fell to 90 — you bought more, and the average is 95. It seems that break-even is now closer. But the position has doubled, and if the price keeps falling to 80, the loss is 30, while a stop at 90 on the original position would have limited it to 10. Averaging down turns one unsuccessful trade into a big one. You may add to a position under a rule written down in advance, but not because the position is in the red.

3.Moving the stop further away

The price has come close to the stop, and the stop is moved “just a little further”. This is the same as trading without a stop, only with self-deception. A stop is moved only in the direction that reduces risk.

4.Too large a size and too much leverage

With a risk of 10% of capital per trade, five losses in a row take about 41% of the account, and to get back to where you started you need to earn about 70%. With a risk of 1%, the same five losses take about 5%. The leverage you are allowed is not the leverage you are advised to use (lesson Cross and isolated margin).

5.Overtrading

Every trade pays costs. Twenty round trips a day at 7 bp on stocks come to 1.4% of the position a day on fees and the spread alone. Unnecessary trades — outside the plan, out of boredom or excitement — add costs and mistakes (lesson Execution and trade cost). Frequent trades are a normal part of scalping when the round trip is cheap and every trade follows the plan.

6.Revenge trading after a loss

After a loss you want to get the money back right away — with a bigger position, outside the plan. This is exactly how one unsuccessful trade turns into an unsuccessful day. A daily loss limit is written into the plan in advance, and trading stops once it is reached (lesson Trading plan and journal).

7.Belief in a “sure-fire” signal

A footprint, tape or indicator signal is a hint, not a promise of a move. A single sign is a reason to take a closer look; read it together with the level, the market regime and the risk. Claims like “after this signal the price always reverses”, made without data and a methodology, are a reason for doubt.

8.Changing the rules after a few trades

Even rules with a positive expectancy have long losing streaks. Below are trades with a 40% win rate and wins twice the size of the losses: the expectancy is positive, but look at the losing streaks and drawdowns that occur.

Series of trades

First 30 series: cumulative result in R; dashed — the expectancy

Expectancy per trade
+0.20 R
Expectancy over 100 trades
+20.0 R
Series ending below zero
9.8 %
Series result: 5% · median · 95%
−4.0 · +20.0 · +44.0 R
Largest drawdown: median · 95%
10.0 · 19.0 R
Losses in a row: median · 95% · max
7 · 13 · 19

500 series; trades are independent, a win is the given number of R, a loss is 1 R. In real trading results are related, and losing streaks are usually longer.

Whoever changes the rules after five losses in a row will never find out whether they worked. Conclusions are drawn from hundreds of trades (lesson Trading plan and journal).

9.Overfitting

If you try many variants of a rule on the same data, it is easy to find the “best” one — even if all the variants are random. Below is the best of 20 variants that have no edge at all:

The best of random strategies

Thin lines — all variants (up to 60), bold — the best in history, dashed — the same on new data. Every variant has an expectancy of exactly zero.

The best variant in history
+28.4 R
Its t-statistic
+2.09
Winning trades
59.0 %
Variants with t > 2
1 / 20
The same variant on new data
−19.8 R
t on new data
−1.41

Each variant is a series of trades with results N(0; 1) in R: none has an edge. Choosing the best by history is choosing the luckiest case, not the best rule.

The best of the random variants looks convincing. The defense is to test the rules on data that played no part in choosing them, and on new trades.

10.No records

Without a journal you cannot tell bad rules from bad execution, and memory obligingly keeps the successful trades. A record of every trade is the minimum without which no review is possible.

In the JetScalp terminal

A stop loss can be placed automatically when a position is opened — the Auto stop loss block in the order book settings (Stop loss). Your results by trade, without relying on memory, are in the Statistics window.

Summary

  • A stop is placed before the entry and moved only in the direction that reduces risk; the position size is calculated from the stop.
  • Averaging down and revenge trading turn one unsuccessful trade into a big loss.
  • Costs make overtrading expensive; a large risk per trade makes a losing streak destructive.
  • Good rules also have losing streaks, and random ones also have a best of many variants — conclusions are drawn from hundreds of trades and on new data.

Check yourself

  1. Why, after losing 50% of the account, do you need to earn 100% to get back to where you started?
  2. How does adding to a position under a rule differ from averaging down?
  3. Why can’t the best of 20 rule variants on the same data be considered tested?
Last updated on