PLATFORM
VOL SURFACE·SVI parametric · NIFTY + BANKNIFTY + F&O
EQUITY RESEARCH·DCF · Comps · Football Field
IPO PIPELINE·GMP feed · 5-broker bid placement
SUPPLY CHAIN·Upstream / downstream graph · NSE
TRADE JOURNAL·Greeks attribution · pattern detection
STRESS TESTING·COVID · 2008 · Taper · custom shocks
MONTE CARLO VAR·10k paths · Cholesky correlated
OPTIONS GREEKS·Δ · Γ · Θ · Vega · Rho — full chain
ARTHA AI·multilingual · cited sources · deep mode
DIVIDENDS·Forward calendar · 200+ Indian ex-dates
MPC ANALYSIS·RBI rate decisions · regime detection
5 BROKERS·Zerodha · Dhan · Upstox · Angel · Fyers
FACTOR MODEL·Fama-French 5 + MOM · alpha isolation
ROLLING CURVES·β · Sharpe · vol · correlation · daily
ROOT-CAUSE·Mistake taxonomy · emotion × outcome
VOL SURFACE·SVI parametric · NIFTY + BANKNIFTY + F&O
EQUITY RESEARCH·DCF · Comps · Football Field
IPO PIPELINE·GMP feed · 5-broker bid placement
SUPPLY CHAIN·Upstream / downstream graph · NSE
TRADE JOURNAL·Greeks attribution · pattern detection
STRESS TESTING·COVID · 2008 · Taper · custom shocks
MONTE CARLO VAR·10k paths · Cholesky correlated
OPTIONS GREEKS·Δ · Γ · Θ · Vega · Rho — full chain
ARTHA AI·multilingual · cited sources · deep mode
DIVIDENDS·Forward calendar · 200+ Indian ex-dates
MPC ANALYSIS·RBI rate decisions · regime detection
5 BROKERS·Zerodha · Dhan · Upstox · Angel · Fyers
FACTOR MODEL·Fama-French 5 + MOM · alpha isolation
ROLLING CURVES·β · Sharpe · vol · correlation · daily
ROOT-CAUSE·Mistake taxonomy · emotion × outcome
// Portfolio Intelligence

Know why your P&L looks the way it does.

Most platforms show you what happened. Kautilya shows you why. Factor regression isolates your alpha from market tilts. Rolling curves track your risk through time. The root-cause engine dissects every losing trade — sizing, timing, emotions, plan adherence — so you stop repeating the same mistakes.

01 · Factor attribution

Your return, decomposed.

You made 25% last year. How much was the market? How much was your value tilt? How much was actual alpha? Most retail investors can't answer. Kautilya runs a Fama-French 5-factor + momentum regression on your portfolio — weekly returns, 2-year window, India-calibrated — and breaks your return into six factor betas plus a residual alpha.

If R² is 0.91, factor loadings explain 91% of your return variation. The remaining 9% is your genuine stock-picking edge (or lack of it). Every beta comes with a t-statistic — green means statistically significant at 95%. This is the same decomposition institutional allocators run on their PMS mandates.

Factor Attribution · Sample ₹25L Portfolio
Fama-French 5 + Momentum · 2-year weekly regression
0.91
Alpha (ann.)
+4.2%
Factor Return
+18.6%
Specific Return
+6.8%
Market (β)
+62.1%
β = 0.87
HML (Value)
+14.2%
β = 0.24
SMB (Size)
-4.8%
β = -0.12
MOM (Momentum)
+18.6%
β = 0.31
QMJ (Quality)
+8.4%
β = 0.18
BAB (Low-Vol)
+3.2%
β = 0.09
Market
t = 12.4
HML
t = 3.8
SMB
t = -1.9
MOM
t = 4.2
QMJ
t = 2.6
BAB
t = 1.4
02 · Rolling analytics

Risk isn't static. Watch it move.

A single-number beta is a snapshot. A rolling beta is a movie. Kautilya computes 30-day, 90-day, and 180-day rolling windows for beta, Sharpe, volatility, alpha, and inter-holding correlations — updated daily. When your 30-day beta spikes from 0.85 to 1.3, you see it before the next drawdown proves it.

The sparklines below are the same rolling curves compressed to glanceable thumbnails. In the workstation, each is an interactive chart with hover crosshairs, regime shading, and the ability to overlay any two metrics.

Rolling Beta vs NIFTY 50
2-year lookback · daily returns
30-day90-day
0.60.81.01.21.4β=1Jun 24Sep 24Dec 24Mar 25Jun 25
Current β (30d)
0.96
Current β (90d)
0.84
Rolling Sharpe
1.38
Rolling Vol (ann.)
16.2%
Rolling Sharpe (90d)
1.26
Rolling Vol (90d ann.)
15.2%
Rolling Alpha (90d)
+4.3%
03 · Trade root-cause engine

Stop losing the same way twice.

The journal tells you what happened. The root-cause engine tells you why. It categorises every losing trade into a mistake taxonomy — sizing, timing, symbol selection, hold duration, Greeks underestimation — and quantifies the damage. Then it correlates your behavioral flags (revenge, FOMO, no plan, broken SL) with actual P&L outcomes.

The result: specific, actionable insights. Not "you lost money in March" but "your revenge trades lose 4× more than average, and 80% happen in ELEVATED vol regimes on Tuesdays after 11 AM."

Trades with a plan
62%
148 / 240 trades
SL followed
41%
Avg loss ₹2.1K vs ₹8.4K when broken
TP followed
68%
Avg win ₹4.8K vs ₹2.1K when cut early
Decision score
58/100
Good decisions win 71% — bad ones 23%
Mistake Taxonomy · Last 90 days
Categorised losses by root cause · 84 losing trades analysed
Hold Duration
-₹48.2K34%
Sizing
-₹32.4K23%
Timing
-₹24.8K17%
Symbol Selection
-₹18.6K13%
Greeks Underestimation
-₹12.4K9%
Emotion × Outcome Correlation
How much worse are flagged trades vs your baseline?
Revenge trades12 trades
Avg P&L -₹4,820
Others -₹1,200
WR 18%
Others WR 52%
FOMO entries8 trades
Avg P&L -₹2,140
Others -₹1,200
WR 31%
Others WR 52%
No plan24 trades
Avg P&L -₹1,860
Others -₹820
WR 35%
Others WR 56%
SL broken6 trades
Avg P&L -₹8,400
Others -₹1,200
WR 8%
Others WR 52%
Size escalation5 trades
Avg P&L -₹3,200
Others -₹1,200
WR 22%
Others WR 52%
⚠ Pattern detected

Your revenge trades lose 4× more than average. 12 trades, ₹57K lost — equivalent to 3 months of alpha.

⚠ Pattern detected

Hold duration is your #1 loss driver. You hold losers 2.4× longer than winners — classic disposition effect costing ₹48K.

⚡ Strength identified

You trade well in STABLE vol regimes (62% WR, +₹3.2K avg). Your edge disappears in ELEVATED and inverts in DANGER.

⚡ Strength identified

When you set a stop loss AND follow it, your average loss is ₹2.1K vs ₹8.4K when you don't. The math is clear.

04 · Fair value audit trail

Every assumption, traced to source.

Kautilya's DCF doesn't pull inputs from thin air. Revenue CAGR comes from the annual report. EBITDA margin from 3-year historical average. WACC uses Damodaran's India ERP (11.14%) + CRP (4.12%) + 10-yr G-Sec as risk-free. Every number has a provenance badge — green for annual report, blue for historical, amber for Damodaran default.

The sensitivity matrix lets you stress any two inputs (WACC × terminal growth) and see how fair value shifts. Eight valuation methods — DCF, Reverse DCF, Scenario DCF, RIM, EVA, APV, SOTP, Comps — stacked in a football field. This is the same depth a sell-side analyst delivers, automated for every NSE 500 name.

DCF Assumption Provenance · RELIANCE
Every input traces to a source — no black boxes
Revenue CAGR
Annual Report
14.2%
EBITDA Margin
Historical 3yr
22.8%
WACC
Damodaran India
12.4%
Terminal Growth
GDP proxy
5%
Tax Rate
BSE Filing
25.2%
Intrinsic value: ₹3,060 · LTP ₹2,648 · upside +15.6%

Your P&L has a story. Read it.

Factor decomposition. Rolling risk. Behavioral root-cause. One workstation, zero guesswork.