The asymmetry of recovering losses, diversification and correlation, position sizing by account risk, volatility drag in leveraged and inverse products, and biases such as loss aversion and the disposition effect.
Part 1: Fundamentals of stocks and markets
About 8 min read
As of October 2026 · rules can change
The asymmetry of losses and recovery
Risk management starts with this one table.
Loss
Return needed to get back to even
−10%
+11.1%
−25%
+33.3%
−50%
+100%
−75%
+300%
−90%
+900%
The bigger the loss, the harder recovery becomes — exponentially. So long-term results are often decided less by the best year than by the deepest losing stretch (maximum drawdown, MDD). Avoiding big losses comes before chasing big gains.
The bigger the loss, the more steeply the return needed to recover rises
Return needed to get back to even
Vertical: Return needed · Horizontal: Loss
Volatility eats into returns
Even with the same average yearly return, the one with bigger swings ends up worse. This is called volatility drag.
Same average, bigger swings, worse result (starting with ₩1 million)
+10% → −10%₩990,000 (−1%)
+50% → −50%₩750,000 (−25%)
What diversification can and can’t do: correlation
Spreading money across several stocks reduces the shock one stock’s accident gives the whole. But diversification has limits, because of how much stocks move together — their correlation.
Variance of an equal-weight n-stock portfolio ≈ σ² × (1/n + (1 − 1/n) × ρ)
Risk that shrinks as you add stocks, and risk that remains
Portfolio volatility
Vertical: Annual volatility
The volatility of an equal-weight portfolio when each stock has 30% volatility and the average correlation is 0.3. By about 10 stocks most of the reducible risk is gone, and about 16.4% remains as the risk of the whole market.
Position size: how much to buy
“How much to buy” affects results more than “what to buy”. A common method is to cap the money you can lose on one trade at a set share of the account.
Setting the quantity with the 1% account-risk rule
Account₩10M
Money you can lose at once₩100,000
1% of the account
Risk per share₩4,000
Buy at ₩50,000, stop at ₩46,000
Quantity to buy25 shares
100,000 ÷ 4,000 = ₩1.25 million worth
The traps of leveraged and inverse products
Products that track twice (leveraged) or the opposite (inverse) of an index’s daily return are reset every day. Held over several days, the result is not “twice the return for the period”.
If the index alternates +5% and −5% every day for 20 days
Index
2× leveraged
Vertical: Start = 100 · Horizontal: Day
The index fell about 2.5%, but the 2× leveraged product fell about 10%. In a market that moves up and down without direction, leveraged products keep eroding.
How to design stop-loss rules
Tie it to your reason: “sell if the reason I bought (e.g. the support level holding) breaks” is more consistent than “sell at −7%”. You can practice levels such as below support or a trendline break in Support and resistance and Trendlines.
A width that fits the volatility: put a 2% stop on a stock that usually moves 3% a day and you’ll be stopped out by every wobble.
Gaps jump over stop prices: if an earnings release or overnight event makes the stock open far below your stop, the actual loss is bigger than planned. That is why you may need to adjust your position before earnings.
Write the rule down: noting the stop level when you buy reduces the temptation to move it when the price actually gets there.
Behavioral biases: habits that fool us
Bias
What it looks like
Response
Loss aversion
The pain of losing feels about twice as strong as the joy of gaining the same amount
Set your loss limit as a number in advance
Disposition effect
Selling winners quickly and holding losers until they “break even”
Judge by the reason to hold now, not by your purchase price
Confirmation bias
Looking only for information that supports your view
Write down in advance “what signal would show I’m wrong”
Anchoring
Judging cheap or expensive by an old high or your purchase price
Judge again on value and current information
Overconfidence
Treating a few wins as skill and increasing position sizes
Check your win rate and payoff ratio in numbers with a trading journal
Recency bias
Believing the trend of the last few months will continue
Compare with longer-period charts (weekly, monthly)
Expectancy: what matters more than win rate
You can’t judge a trading method by its win rate alone. You have to look at how much you make when you win and how much you lose when you lose.
Expectancy per trade = win rate × average gain − (1 − win rate) × average loss
Method
Win rate
Average gain
Average loss
Expectancy (per trade)
A
70%
+3%
−8%
0.7×3 − 0.3×8 = −0.3%
B
40%
+9%
−4%
0.4×9 − 0.6×4 = +1.2%
A, with a 70% win rate, wins often but loses in the end; B, with a 40% win rate, loses often but earns in the end. The habit of grabbing small gains quickly and sitting on losses (the disposition effect) produces results like A. Subtract trading costs (chapter 3) to get the real expectancy.
A high win rate can still lose: expectancy per trade
Method A70% wins, +3% / −8%−0.3%
Method B40% wins, +9% / −4%+1.2%
Rebalancing: rules for restoring weights
Even if you start with 60% stocks and 40% cash and bonds, the stock weight grows by itself when prices rise and shrinks when they fall. Periodically restoring the weights you set is called rebalancing.
Keeping risk steady: as the weight of assets that rose grows, you carry more risk than you first chose. Rebalancing brings risk back to its original level.
The benefit of a rule: you end up selling some of what rose and buying what fell, which makes you act against your emotions.
Method: do it on set dates (quarterly, half-yearly) or when weights drift beyond a set band (e.g. ±5 percentage points). Doing it too often piles up costs and taxes.
Before and after rebalancing (restoring 60% stocks, 40% cash)
Before rebalancing
Stocks ₩9M (69%) · cash ₩4M
After rebalancing
Stocks ₩7.8M (60%) · cash ₩5.2M
Stocks
Cash
With stocks up 50% to a 69% weight, restoring 60% cuts the loss from a later 30% fall in stocks from ₩2.7 million to ₩2.34 million.
This chapter reflects the situation as of October 2026. Trading rules, tax rates and fees can change, so check with the exchange, the tax authority or your broker before you trade or file. Parts marked “to be confirmed” are especially likely to change. This is learning material and does not recommend any stock or trade.