How to Trade — Execution & Platforms

Q1 (12)

(a) Derivation (4)

• Rupee risk budget $= r \cdot E$. (why: policy caps loss to a fraction of equity) (1)
• Loss per share if stop hit $= |P_e - P_s|$. (why: this is the per-unit adverse move you accept) (1)
• Shares × per-share loss $=$ total loss allowed $\Rightarrow N \cdot |P_e-P_s| = r E$. (1)
• $N = \frac{r \cdot E}{|P_e - P_s|}$ Assumptions: stop is honoured at $P_s$ (no slippage/gap), no leverage constraint binding. (1)

(b) Computation (5)

• Risk per share $= 250 - 238 = ₹12$. (1)
• Rupee risk budget $= 0.015 \times 5{,}00{,}000 = ₹7{,}500$. (2)
• $N = 7500 / 12 = 625$ shares (already integer). (2)

(c) Explain (3) — Fixed capital allocation ignores volatility: a wide-stop trade would then risk far more rupees than a tight-stop trade. Tying size to stop distance equalises rupee risk across trades, so no single loss exceeds the policy cap. (3)

Q2 (11)

(a) Derivation (4)

• $L = \text{PV}/M$. (1)
• P&L $= \text{PV}\cdot m$ where $m$ = % move (decimal). (1)
• Return on margin $R = \text{P\&L}/M = \text{PV}\cdot m / M = L \cdot m$. (1)
• $R = L \cdot m$ (why: leverage multiplies the underlying % move onto your margin) (1)

(b) (4)

• $L = 200000/40000 = 5\times$. (1)
• P&L $= 200000 \times 0.03 = ₹6{,}000$. (2)
• $R = 6000/40000 = 15\% = 5 \times 3\%$. ✓ (1)

(c) (3)

• $R = 5 \times (-3\%) = -15\%$; rupee loss $= -₹6{,}000$. (2)
• Lesson: leverage amplifies gains and losses identically; a $3\%$ adverse move wipes $15\%$ of margin. (1)

Q3 (10)

(a) Derivation (6)

• Let price fall by fraction $d$. New position value $= 100000(1-d)$. (1)
• Debt stays $₹50{,}000$; equity $= 100000(1-d) - 50000$. (1)
• Margin call when equity $<$ maintenance $= 0.30 \times \text{position value}$: $100000(1-d) - 50000 < 0.30\cdot100000(1-d)$. (1)
• $0.70\cdot100000(1-d) < 50000 \Rightarrow 70000(1-d) < 50000$. (1)
• $1-d < 5/7 \Rightarrow d > 2/7 \approx 0.2857$. (1)
• Margin call triggers at ~28.57% drop (price $\approx ₹71.43$ per ₹100 of original). (1)

(b) Explain (4) — Auto square-off = broker automatically closes your position when equity breaches maintenance and you don't top up. (2) Enforced because the broker's borrowed money is at stake; closing before equity hits zero protects the broker from a negative-balance (unrecoverable) account. (2)

Q4 (9)

(a) Explain (4) — SEBI takes 4 random intraday snapshots of client margin; the peak (highest requirement across snapshots) must have been collected upfront, not just end-of-day. (2) This killed the old model where brokers gave huge intraday leverage and only checked margin at day-end; now full VaR+ELM upfront caps leverage sharply. (2)

(b) (5)

• Old $20\times$: max position $= 25000 \times 20 = ₹5{,}00{,}000$. (2)
• New $5\times$: max position $= 25000 \times 5 = ₹1{,}25{,}000$. (2)
• Reduction factor $= 500000/125000 = 4\times$ smaller (75% cut). (1)

Q5 (10)

(a) (3)

• Spread $= 100.20 - 100.00 = ₹0.20$. (1)
• Mid $= 100.10$. (1)
• bps $= (0.20/100.10)\times 10000 \approx 19.98$ bps $\approx 20$ bps. (1)

(b) (5)

• Spread cost on 1000 shares $= 0.20 \times 1000 = ₹200$. (2)
• Brokerage both sides $= 2\times20 = ₹40$; taxes $= ₹35$. (1)
• Total cost $= 200 + 40 + 35 = ₹275$. (1)
• Break-even move per share $= 275/1000 = ₹0.275$. (1)

(c) (2) — A $0.10$/share edge earns $₹100$ on 1000 shares but the spread+cost is $₹275$; net $-₹175$ per round trip. The strategy has negative expectancy purely from friction. (2)

Q6 (8)

(a) (4) — Any 4 sensible bindings, ~1 mark each: e.g. F1 = buy market, F2 = sell market, F3 = flatten/close all, Esc = cancel all orders, Shift+B = buy at bid, arrow keys = adjust price. Justification: keeps hands on keyboard, removes mouse-hunt latency in fast markets, one-key exit reduces panic errors. (4)

(b) (4) — Alerts convert "watching" into "waiting": set price/indicator alerts so the platform pings only when a setup is live, letting you monitor 10+ symbols passively. (2) Example: "Alert when RRR crosses above 20-day high on volume > avg" fires a push notification; trader then acts only on that symbol instead of scanning all charts. (2)

[
  {"claim":"Q1b position size = 625 shares","code":"r=Rational(15,1000); E=500000; Pe=250; Ps=238; N=(r*E)/(Pe-Ps); result = (N==625)"},
  {"claim":"Q2b return on margin = 15%","code":"PV=200000; M=40000; m=Rational(3,100); R=PV*m/M; result = (R==Rational(15,100))"},
  {"claim":"Q3a margin call drop = 2/7","code":"d=symbols('d'); sol=solve(Eq(70000*(1-d),50000),d)[0]; result = (sol==Rational(2,7))"},
  {"claim":"Q4b reduction factor = 4","code":"old=25000*20; new=25000*5; result = (old/new==4)"},
  {"claim":"Q5b break-even move = 0.275 per share","code":"cost=0.20*1000+2*20+35; be=cost/1000; result = (abs(be-0.275)<1e-9)"}
]