The short answer
Averaging down means buying more of a trade that is losing, to lower your average price. It makes the break-even price closer, so a small bounce feels like enough to get out.
The catch: every new entry makes the position bigger. So every extra dollar gold falls costs you more than the dollar before. Most of the time gold bounces and you escape. The one time it doesn’t, the loss is many times what your stop would have been.
It is the everyday version of a martingale. Our martingale page tested the full doubling system on real gold data; this page shows the version most traders do without giving it a name.
The worked example (all numbers are examples)
The maths was checked by script.
- Account: $5,000.
- Plan: buy 0.10 lot at $4,000, stop at $3,990. Risk: $10 × 10 oz = $100 = 2% of the account.
Gold falls to the stop. Instead of taking the $100 loss, the trader cancels the stop and adds 0.10 lot every $10:
| Entry | Lots | Total lots | Average price |
|---|---|---|---|
| $4,000 | 0.10 | 0.10 | $4,000 |
| $3,990 | 0.10 | 0.20 | $3,995 |
| $3,980 | 0.10 | 0.30 | $3,990 |
| $3,970 | 0.10 | 0.40 | $3,985 |
If gold bounces
Gold comes back to $3,995. The 0.40 lot position is $10 above its average: +$400. The trader feels clever. The original trade was a loser, and now it’s a winner. This is the trap: the habit is rewarded.
If gold keeps falling
| Gold price | Loss on 0.40 lot (average $3,985) | Share of the account |
|---|---|---|
| $3,960 | −$1,000 | 20% |
| $3,940 | −$1,800 | 36% |
The planned loss was $100. At $3,940 it is $1,800, eighteen times bigger, and gold is only $60 below the first entry.
Add one more 0.10 lot at $3,960 “to get a better average” (0.50 lot, average $3,980), and at $3,940 the loss is −$2,000 = 40% of the account.
The margin side
At 1:100 leverage (example), the five entries use about $1,990 of margin. With equity at $3,000 after the −$2,000 loss, the margin level is $3,000 ÷ $1,990 ≈ 151%. A few more dollars down and the margin call warnings start. The more you add, the sooner the stop-out comes.
The doubling version (everyday martingale)
Some traders double instead: 0.10 lot at $4,000, 0.20 at $3,990, 0.40 at $3,980.
- Total: 0.70 lot, average price about $3,985.71.
- At $3,960, the loss is −$1,800 = 36%, after only a $40 fall from the first entry.
- The bounce needed to break even is $25.71, about the same as the equal-size version. But the loss if it doesn’t bounce grows much faster.
Doubling makes the “small bounce” feel even closer. It also makes the account fall off a cliff sooner. Our martingale simulations on real gold data show what happens over years, and the grid and hedging page covers the robot versions.
Why it feels so right
- A lower average price feels like progress. The break-even line moves closer with each entry.
- Taking a loss feels worse than risking a bigger one. Psychologists call this loss aversion. It is the same force behind holding losers too long.
- It works often. Gold does bounce after many dips, so the habit gets rewarded many times before it gets punished once.
- “It’s cheaper now.” That is how investors think about gold coins. A leveraged trader’s position is not a coin: it has margin, a stop-out, and a balance that can run out.
Averaging down vs a planned scale-in
Adding to a position is not always wrong. The difference is when you decide.
| Averaging down | Planned scale-in | |
|---|---|---|
| Decided | After the trade goes wrong | Before the first entry |
| Stop | Removed or moved | One stop for all entries, set at the start |
| Total risk | Grows with every add | Fixed, e.g. 1% for all entries together |
| Trigger | “It’s cheaper” | A written rule, e.g. “half at the level, half on the retest” |
Example of a planned scale-in (example numbers): $5,000 account, 1% total risk = $50. Stop at $3,990 for everything. Entry 1: 0.02 lot at $4,010 (risk $40). Entry 2: 0.01 lot at $4,000 (risk $10). Total risk if the stop is hit: $50, decided in advance. That is a sizing method, not a rescue.
Adding to winners (pyramiding) is another planned method: you add only after the trade moves in your favour and the stop on the whole position is raised so the total risk stays within your limit.
How to stop averaging down
- Stop in the platform before you click. A trade with a stop can’t be “rescued” without a deliberate choice. See moving your stop-loss.
- Write the maximum size for the trade on your rules card: “Max 0.10 lot on gold. No new entries below my entry price.”
- Use the position size calculator for every entry, including planned scale-ins: position size calculator.
- When you feel the urge to add, step away for ten minutes. The urge is strongest right after a fast move against you.
- Look at the recovery maths. A 36% loss needs a 56% gain to get back. Try it in the drawdown recovery calculator.
Warning signs you’re doing it
- You have three or more positions on gold in the same direction, all opened at worse prices.
- You removed or widened the stop “just this once”.
- You talk about the trade’s “average” instead of its risk.
- You feel relief, not pride, when it closes.
If this has become a pattern that you can’t stop, our page on when trading becomes gambling has warning signs and real help lines.
Common mistakes
- Calling a rescue a “strategy”. If the extra entries weren’t in your plan before the first click, it’s a rescue.
- Adding because of margin space. Free margin is not spare risk.
- Holding an averaged-down position over the weekend, where a gap can skip past every level you were counting on.
- Telling yourself gold “always comes back”. Gold has had long falls: see why gold fell in 2026.
Sources
- All prices, lot sizes, leverage and the account are example numbers. The maths (average price, losses, margin level, doubling version, recovery percentage) was checked by script.
- Loss aversion: Kahneman, D. and Tversky, A. (1979), “Prospect Theory: An Analysis of Decision under Risk”, Econometrica 47(2), 263–291, doi:10.2307/1914185.
Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.
Frequently asked questions
Is averaging down a good strategy for gold?
For a leveraged trader it is one of the fastest ways to lose an account. Each extra entry lowers your average price but raises your position size, so the loss grows faster with every dollar gold falls. It often works for a while, which hides the risk until one move doesn’t come back.
What is the difference between averaging down and martingale?
Martingale doubles the size after each loss or each step against you. Averaging down adds to the losing position, often with equal or growing size. Both increase exposure when you are wrong, and both depend on price coming back before the account runs out.
Is it ever OK to add to a gold position?
Adding to a winning trade, or a planned scale-in where every entry and one shared stop are decided before the first click and the total risk stays within your limit, is a different thing. What causes the damage is adding unplanned size to a trade that is already losing.
Why does averaging down feel so right?
Because it lowers the break-even price, so a small bounce feels like enough. It also delays admitting the trade was wrong. Loss aversion makes taking a loss feel worse than risking a bigger one.
How do I stop averaging down?
Set the stop and the maximum position size before you enter, put both in the platform, and write a rule: no new entries on a trade that is below its entry price. If you feel the urge, close the platform for ten minutes.
Comments
No comments yet.