Understand scaling in and scaling out

What exactly is scaling?

WHAT problem does it solve?

Timing uncertainty — you rarely nail the exact bottom/top.
Emotional pressure — a giant all-in trade makes you panic; small pieces feel manageable.
Reward-vs-regret — scaling out captures profit and keeps upside, reducing "I sold too early" and "I gave it all back" regret simultaneously.

Deriving position math from first principles

Let total shares $= N$ , split into $k$ tranches. Tranche $i$ has size $n_i$ and fill price $p_i$ , so $N=\sum_{i=1}^{k} n_i$ .

Worked Example 1 — Scaling IN (pyramiding a winner)

You plan to buy 300 shares of a stock breaking out at ₹100. You scale in:

Tranche	Shares	Price
1	100	₹100
2	100	₹104
3	100	₹110

Step 1 — total spent. $100(100)+100(104)+100(110)=10000+10400+11000=₹31400$ . Why? Cost basis needs total money, so sum each tranche's cost.

Step 2 — average price. $\bar p = 31400/300 = ₹104.67$ . Why? Adding higher raises your average — that's the trade-off for confirmation.

Step 3 — price now ₹112, P&L. $300(112-104.67)=300(7.33)=₹2199$ . Why? Every share earns $s-\bar p$ .

Worked Example 2 — Scaling OUT

You own 300 shares, average cost $\bar p=₹104.67$ , stop at ₹98. You sell in thirds:

Sell	Shares	Price
1	100	₹110
2	100	₹120
3 (runner)	100	₹135

Step 1 — P&L. $100(110-104.67)+100(120-104.67)+100(135-104.67)$ $=533+1533+3033=₹5099$ . Why? Sum $(s_j-\bar p)$ over each slice.

Step 2 — average exit. $(110+120+135)/3=₹121.67$ (equal sizes). Why? Weighted mean collapses to simple mean when tranche sizes are equal.

Step 3 — R-multiple. $R=104.67-98=6.67$ . Avg gain per share $=121.67-104.67=17$ . So $17/6.67 = 2.55\text{R}$ . Why? You made ~2.5× the money you risked — the real scorecard.

Figure — Understand scaling in and scaling out

Common mistakes (steel-manned)

Imagine buying candy for a party but you're not sure how many friends will show up. Instead of buying ALL the candy at once, you buy a little first. If more friends arrive (price goes your way), you buy more. When you have loads, you start giving it away in handfuls — some now, some later — so you always have a little left in case something great happens, but you never risk running out completely. That "buy-a-bit, sell-a-bit" is scaling.

Active Recall

What is scaling in?

Building a position in multiple partial buys (tranches) instead of one order, typically adding as the trade confirms.

What is scaling out?

Closing a position in multiple partial sells to lock profit while letting a remaining portion run.

Formula for average entry price of k tranches?

\bar p=\frac{\sum n_i p_i}{\sum n_i}

— total money divided by total shares (weighted mean).

Why scale IN higher rather than average down?

Adding higher only increases size when you're already right; averaging down grows the position as the thesis fails.

What is the hidden cost of scaling OUT in a strong trend?

Lower total profit than holding the full position — you trade expected value for reduced variance/regret.

Define R-multiple.

Profit per share divided by initial risk per share:

(s-\bar p)/(\bar p-\text{stop})

How do you keep total risk fixed while scaling in?

Use one stop for the whole position and size later tranches smaller so total "heat" stays within your % limit.

A tranche is?

One slice/partial order of the total intended position size.

Connections

Position Sizing — scaling only works if total heat stays capped.
Stop-Loss Placement — the stop defines $R$ used in R-multiples.
Pyramiding — the disciplined form of scaling in.
Risk-Reward Ratio — scaling out reshapes your realized R distribution.
Trade Management — parent topic.
Trend Following — where letting a runner pays most.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, scaling ka matlab simple hai: apni poori position ek saath mat kharido ya becho, tukdon mein karo. Scaling in ka matlab — trade jab tumhe sahi prove karti jaaye (price upar jaaye), tab thoda-thoda add karte jao. Ise pyramiding kehte hain. Aur scaling out ka matlab — profit book karne ke liye thoda-thoda bechte jao, lekin thodi quantity "runner" ke roop mein chhod do jo bade move mein extra profit de.

Yeh important kyun hai? Kyunki future kisi ko nahi pata. Agar tum ek hi baar full paisa laga do, toh tumhe direction aur exact timing dono perfect chahiye — bahut mushkil. Scaling se tum "lagbhag sahi" hoke bhi paisa bana lete ho, aur risk chhota rehta hai jab tak market confirm na kare. Average entry price bas ek weighted average hai: total paisa divided by total shares — koi jaadu nahi.

Ek badi galti se bacho: averaging down ko scaling in mat samjho. Girte hue stock mein aur kharidna "sasta mil raha hai" feel deta hai, par asal mein tum galat trade mein size badha rahe ho — risk bloat ho jaata hai. Discipline yeh hai: add tabhi karo jab thesis abhi bhi valid ho, aur baad ke tranches chhote rakho taaki total risk (heat) tumhari fixed % limit ke andar rahe.

Aur yaad rakho — scaling out strong trend mein tumhara total profit kam kar deta hai full-hold ke comparison mein. Yeh profit maximize karne ka tool nahi, balki emotion aur variance control karne ka tool hai. Mnemonic: "IN as it earns, OUT as it churns."

What exactly is scaling?

Deriving position math from first principles

Worked Example 1 — Scaling IN (pyramiding a winner)

Worked Example 2 — Scaling OUT

Common mistakes (steel-manned)

Feynman

Active Recall

Connections

Concept Map

Hinglish (regional understanding)

Test yourself — Entry, Exit & Trade Management

Connections