Last updated: 1 August 2026 · 21 min read · The definitive guide
By Pulkit Mangal — equity + F&O trader since 2017, founder of TradeDiary. Traded across Zerodha, Kotak, Upstox and Dhan; built TradeDiary after losing ₹14L in FY21 to mistakes that were sitting in plain sight in my own trade data.
TL;DR: Trading analytics is the practice of turning your own closed trades into a small number of decision-grade statistics — win rate, R-multiple, expectancy, and drawdown — and then slicing those by setup, instrument and time so you can tell which parts of your trading make money and which quietly destroy it. For an Indian trader it only works if three things are true: P&L is measured net of STT, brokerage, GST and stamp, equity quantity is matched FIFO the way the Income Tax Act requires, and F&O is grouped by each underlying's native expiry cycle rather than the calendar week. Everything below is the full stack, in the order you should build it. If you read one line: expectancy is the only number that tells you whether you have an edge — and it is meaningless computed on gross P&L.
1. What trading analytics actually is
Trading analytics is statistics computed on your own closed trades, for the purpose of changing what you do next. That last clause is the whole definition. A number that cannot change a decision is decoration.
It sits one layer above journaling. A trading journal is the record — the ledger of what you traded and the one-sentence note on why. Analytics is what you do to that record: aggregate it, slice it, and reduce it to four or five statistics you can act on. Journal without analytics and you have a diary. Analytics without a journal and you have your broker's P&L statement, which tells you the score but never the cause.
Three levels, and most traders stop at the first:
- Descriptive — what happened. Net P&L, number of trades, win rate. Your broker gives you most of this.
- Diagnostic — why it happened. P&L by setup tag, by instrument, by time-of-day, by holding period. This is where the money is, and no broker computes it, because your broker doesn't know your setups.
- Prescriptive — what to do about it. "Stop taking C-grade setups after 14:15" is a prescription. It comes out of the diagnostic layer, and it's the only output that changes your P&L.
The trap is thinking analytics means more numbers. It means fewer numbers, computed correctly, sliced along dimensions that map to decisions you can actually make. Forty columns of statistics is not analytics; it's a way of feeling rigorous while changing nothing.
2. Why Indian traders need it specifically
Most of the analytics writing you'll find was built for a US equity trader paying zero commission on a market with no STT and no FIFO mandate. Three structural facts break that advice at the border.
Costs are itemised, per-leg, and large enough to invert a result. STT on equity delivery is 0.1% per side; on options it's 0.15% on sell-side premium (since 1 April 2026). Add ₹20/order brokerage on discount brokers, exchange transaction charges, SEBI turnover fees, 18% GST on brokerage plus transaction charges, and stamp duty on the buy side. For a high-frequency intraday or scalping style, costs routinely consume 30–60% of gross edge. Analytics computed on gross P&L will confidently tell you a losing strategy is profitable. This is the single most common analytical error I see in Indian traders' spreadsheets.
FIFO is not optional. The Income Tax Department matches equity cost of acquisition on a First-In-First-Out basis. If your analytics uses average price — which is what almost every homemade spreadsheet does, because it's easier — then your per-trade P&L, your holding-period buckets, and your short-term/long-term split are all wrong. They'll be directionally fine and specifically wrong, which is the worst kind of wrong for tax.
Expiry cycles are the real unit of time in F&O. NIFTY, BANKNIFTY, FINNIFTY and MIDCPNIFTY have different expiry days, and the payoff behaviour of an options position depends far more on days-to-expiry than on which calendar week it happened to fall in. Bucketing options P&L by calendar week smears expiry-day behaviour — the single most distinctive thing about Indian retail F&O — into noise.
And the reason any of this matters: as per SEBI's January 2024 study, 9 out of 10 individual F&O traders lost money over FY22, with average losses around ₹1.1 lakh. The winners and losers traded the same instruments on the same screens. What separated them was behaviour, and behaviour is only fixable once it's counted. (SEBI study, Income Tax Act §44AB)
💡 Want your own numbers before you read 4,000 words? Create a free TradeDiary account and import your last 50 trades — no card needed. Your expectancy and drawdown will be computed for you in about two minutes. Come back to the guide once you've seen them.
3. Original data
We work with roughly 500 Indian retail traders and review anonymised, aggregated patterns from their analytics — no personal data, no live positions, just shape-of-behaviour statistics. Treat what follows as directional observations from our user base, not a market-wide study.
Where the edge actually lives — net expectancy by setup grade (R per trade, FY26, equity intraday + index options):
A-grade (planned, in playbook) ████████████████████ +0.34 R
B-grade (planned, marginal) ████████░░░░░░░░░░░░ +0.11 R
C-grade (unplanned, "looked ok") ██████░░░░░░░░░░░░░░ −0.19 R
Revenge / tilt (post-loss) ████████████████░░░░ −0.52 R
Net-of-cost expectancy per trade by self-assigned setup grade, across ~500 anonymised journals. Bar length is magnitude, not sign.
The striking part isn't that unplanned trades lose. It's the ratio: the median trader in our sample took 2.3 C-grade and tilt trades for every A-grade one. The edge was real and it was being funded straight into the leak. No new strategy was needed — only the count.
The analytics gap — what traders think they measure vs what holds up (share of sampled journals):
| Metric | Traders who claim to track it | Tracking it correctly | The usual error |
|---|---|---|---|
| Win rate | 94% | 91% | Fine — but over-weighted in decisions |
| Net P&L (after all charges) | 63% | 28% | Gross P&L, or charges estimated once and reused |
| R-multiple | 41% | 19% | Risk taken as actual loss, not planned risk at entry |
| Expectancy | 22% | 9% | Computed on gross, or on fewer than 20 trades |
| Max drawdown | 31% | 14% | Measured on realised P&L only, ignoring open positions |
| P&L by setup tag | 37% | 22% | Tags added retrospectively (hindsight-contaminated) |
The pattern: the further down the list, the more decision-useful the metric — and the less reliably it's computed. Win rate is tracked almost universally and is the least actionable number here. Expectancy is tracked by one in five and is the only one that answers "do I have an edge."
4. The metric stack
Build these in order. Each one depends on the one above it being correct, and skipping ahead produces confident nonsense.
| # | Layer | What it answers | Depends on |
|---|---|---|---|
| 1 | Net P&L per trade | Did this trade actually make money? | Itemised STT/brokerage/GST/stamp, FIFO matching |
| 2 | Planned risk (R) | How much was I willing to lose at entry? | Stop-loss logged before the outcome |
| 3 | R-multiple | How big was this result relative to what I risked? | Layers 1 + 2 |
| 4 | Win rate & avg win/loss R | What's the shape of my results? | Layer 3, ≥30 trades |
| 5 | Expectancy | Do I have an edge at all? | Layer 4 |
| 6 | Drawdown / risk of ruin | Can I survive long enough to realise it? | Layer 1, equity curve |
| 7 | Sliced expectancy (by tag, instrument, time) | Where does the edge live? | Layers 5 + consistent tagging |
Layer 7 is the payoff. Layers 1–6 exist to make layer 7 trustworthy. And layer 2 is where most traders silently break the chain: if you record risk as the loss you actually took rather than the stop you set at entry, every R-multiple downstream is retrofitted to the outcome and your analytics becomes an elaborate way of agreeing with yourself.
5. Win rate
Win rate is the percentage of your closed trades that finished net-positive. It is the first number every trader computes and the one that misleads most.
A 70% win rate is worthless if your average loss is four times your average win. A 35% win rate is excellent if your winners run 5R. Win rate is one term in a two-term expression, and reading it alone is how traders end up optimising for being right instead of for making money — cutting winners early to protect the percentage, and letting losers run because closing them makes the number look bad. That's not a psychological failing; it's what happens when you measure the wrong thing.
Where win rate is useful: as a stability check within one setup. If your A-grade breakout setup normally wins 48% of the time and this quarter it's winning 29% on 40 trades, something has changed — the regime, your execution, or your entry criteria drifting. Win rate is a good tripwire and a terrible target.
Full treatment, including why the "high win rate = good trader" belief survives: win rate myths every trader believes →
6. R-multiple
R-multiple expresses every result as a multiple of the risk you planned to take at entry. Risk ₹5,000 on a trade and make ₹10,000, that's +2R. Risk ₹5,000 and lose ₹5,000, that's −1R.
This single normalisation is what makes analytics possible at all. Without it, a ₹2,000 profit on a ₹2,000 risk and a ₹2,000 profit on a ₹20,000 risk look identical in your P&L column — and they are ten times apart in quality. Once every trade is in R, you can average them, compare a ₹50,000-capital month against a ₹5,00,000-capital month, and compare an options trade against an equity swing.
Two rules make R honest:
- R is set at entry, before the outcome exists. Entry price minus stop-loss price, times quantity. If you didn't set a stop, you don't have an R — and the fact that you can't compute it is itself the finding.
- R is computed on net P&L. Costs come out first, or your R-multiples are inflated by exactly the amount your broker is taking.
Deep dive plus worked examples: R-multiple explained for Indian traders →. Or compute yours right now with the free R-multiple calculator.
7. Expectancy
Expectancy is the average R you can expect per trade over a large sample:
Expectancy = (Win% × Avg win R) − (Loss% × Avg loss R)
Worked example. You win 42% of trades, average winner +1.9R, average loser −0.85R:
(0.42 × 1.9) − (0.58 × 0.85) = 0.798 − 0.493 = +0.305R per trade
Risk ₹5,000 per trade and that's roughly ₹1,525 expected per trade — before you've said a word about whether the market "feels" right. Take 40 trades a month and the arithmetic says ~₹61,000, with wide variance around it.
Expectancy is the only metric that directly answers do I have an edge. Positive means the system pays; negative means volume makes it worse, not better. Three caveats that matter more in India than anywhere:
- Net or nothing. Compute it on gross P&L and a scalping strategy with +0.05R gross and −0.08R net will read as an edge you should size up. This is how people scale into ruin.
- Sample size. Under ~30 trades per bucket, expectancy is noise wearing a decimal point. Under 20, ignore it entirely.
- Per bucket, not just overall. Aggregate expectancy of +0.1R can hide +0.4R in your A-grade setups and −0.5R in the C-grades. The aggregate is the result; the buckets are the decision.
Full formula walkthrough with Indian cost examples: the trading expectancy formula →. For the wider question of what an edge is and how to prove you have one: what a trading edge actually is →.
8. Drawdown and risk of ruin
Expectancy tells you whether you make money over the long run. Drawdown tells you whether you'll still be trading when the long run arrives.
Max drawdown is the largest peak-to-trough fall in your equity curve, in rupees and in percent. It is a survival metric, not a performance one. A strategy with +0.3R expectancy and a 45% max drawdown is, for most people, unusable — not because the maths fails but because almost nobody keeps executing a system after watching ₹4.5L of ₹10L evaporate. They abandon it two trades before the recovery.
Two mistakes to avoid:
- Measuring on realised P&L only. Your drawdown must include open positions marked to market. A "realised-only" equity curve is smooth by construction and hides exactly the stretch that would have broken you.
- Confusing drawdown with a losing streak. Six losses in a row on 0.5% risk is a 3% dent. One oversized position is a 30% dent. Drawdown is driven far more by position sizing than by being wrong repeatedly.
The related number is risk of ruin — the probability that a run of normal-looking losses takes you below your ability to continue. It's mostly a function of risk-per-trade. At 1% risk with a positive-expectancy system it's negligible; at 10% it's near-certain regardless of edge.
Deep dive: max drawdown in trading — how to measure and survive it →. If you're already in one: how to recover from a trading drawdown →.
9. Slicing the data
Everything so far produces one number for your whole trading. The value comes from cutting it along dimensions that map to decisions.
By setup tag. The highest-yield slice by a distance. Tag every trade with one setup name at entry — orb-breakout, gap-fade, thursday-spread — then compute expectancy per tag. Almost every trader who does this for a quarter finds at least one negative-expectancy tag they've been running for years out of habit. Killing it requires no new skill. Tagging discipline matters: tags applied at entry are data, tags applied afterwards from memory are hindsight. More on schemes that survive contact with reality: trade tagging software and how to build a tag scheme →.
By time-of-day. The Indian session has strong structure — the 09:15–09:45 open-volatility window and the 14:15–15:30 expiry-chase window are where most retail intraday P&L is destroyed. Bucket net P&L into five or six windows and the shape usually appears within 60 trades. Full method: intraday trade analyzer — finding your best and worst hours →.
By instrument and by underlying. Most traders are meaningfully better at one thing than another and don't know which, because P&L is reported as a single blended figure. Split equity from F&O, then split F&O by underlying. NIFTY and BANKNIFTY behave differently enough that blending them hides both.
By holding period. Bucket into intraday / 2–5 days / 2+ weeks. A common and expensive finding: a trader whose swing trades carry the account while intraday activity — the part taking most of the screen time and nearly all the stress — nets out slightly negative after costs.
By position size. Plot R-multiple against size. If your biggest positions have your worst average R, you're sizing up on conviction rather than on edge, which is the most reliable way to make a positive-expectancy system lose money.
10. F&O analytics
Options analytics breaks every naive setup, for one reason: the leg is not the trade. A bull call spread is two legs; analysed as two rows, one leg looks like a disaster and the other like a triumph, and the actual position — the only thing you decided on — appears nowhere. Every leg must roll up to a single strategy-level net P&L before any statistic is computed.
Four requirements specific to Indian F&O:
- Strategy-level aggregation. One tag per position, all legs rolled into one net P&L, one R-multiple. Method: measuring options strategy performance →.
- Expiry-cycle bucketing. Group by days-to-expiry and by each underlying's native cycle, not calendar weeks. A 0-DTE trade and a 12-DTE trade in the same "week" are different businesses.
- Greeks context at entry. Logging entry IV and delta turns "this strategy stopped working" into the answerable "this strategy stops working when IV rank is below 30." That's a filter you can actually apply. See tracking options Greeks in your journal →.
- Assignment and expiry settlement. Positions that expire ITM settle rather than close; if your analytics only counts explicit exits, expired positions silently vanish and your win rate is quietly wrong in your favour.
The India-specific tax note: STT on exercised options is 0.15% of intrinsic value, the same rate as the 0.15% sell-side premium rate since 1 April 2026. If you let ITM options expire rather than squaring off, the cost line in your analytics still needs to include it; it just no longer costs more than closing the position.
Broader context on journaling multi-leg positions: options trading journal for Indian traders →.
11. Behavioural analytics
The SEBI number — 9 in 10 losing — is not explained by strategy selection. The traders in that 90% are not using worse indicators. The gap is behavioural, and behaviour is measurable if you decide to measure it.
Four metrics that quantify discipline:
- Plan adherence rate. Share of trades that were in your playbook before you took them. Below ~70% and no other analytics matters, because you aren't executing the thing you're measuring.
- Post-loss trade interval. Median minutes between a losing exit and the next entry. A cluster under 10 minutes is revenge trading with a timestamp. In our sample it carried a −0.52R expectancy — the psychology behind it, and what's happening neurologically.
- Overtrading ratio. Trades taken vs trades planned, per day. Persistent 2× is boredom or FOMO, not opportunity. Fixes: how to stop overtrading and FOMO trading in Indian markets.
- Stop-loss honour rate. Share of losing trades exited at or before the logged stop. Every miss is an unbudgeted risk and it wrecks your R data as a side effect.
These are the metrics people avoid, because unlike win rate they can't be blamed on the market. That's exactly why they're the ones that move P&L. Track them as one composite with the free discipline score tracker, or read the method: building a trading discipline tracker →. The full psychology stack lives in the companion pillar: the trading psychology handbook →.
12. Costs and tax
Two analytics jobs nobody enjoys and everybody needs.
Cost analytics. Compute cost as a percentage of gross P&L, per strategy. That one ratio reprices your entire book. A scalping strategy showing 55% of gross going to charges isn't a strategy, it's a subscription to your broker. This is also the number that tells you whether trading fewer, larger positions would beat trading more, smaller ones — which for most Indian intraday traders it does, and by a lot.
Tax analytics. Intraday equity is speculative business income; F&O is non-speculative business income; delivery is capital gains with the FIFO cost basis and the STCG/LTCG split. Your analytics has to keep these separate all year, because reconstructing it in July from a broker P&L statement is where errors and panic live. Track F&O turnover as you go — it determines whether the §44AB audit threshold applies, and the definition of turnover for F&O is not your traded value.
The tax angle is covered end-to-end in the journaling pillar: trading journal India — the complete guide →.
13. Tooling
Spreadsheet. Genuinely fine under ~20 trades a month with one broker and no multi-leg F&O. You'll hand-build FIFO matching and a charges formula per instrument type, and you'll maintain them every time a rate changes. The manual entry has real reflective value. It breaks the moment you add a second broker or start trading spreads. Honest comparison: trade journal vs spreadsheet →.
Journal/analytics app. Auto-imports the tradebook, applies FIFO and itemised charges, rolls F&O legs into strategies, and computes the stack in §4 without you maintaining formulas. The tradeoff is a subscription and someone else's opinion about what to measure. What to look for: best trading journal apps in India 2026 → and trading journal apps for Indian traders →.
AI-assisted. Real use: clustering your losing trades by factors you didn't think to tag, flagging behavioural drift, auto-tagging free-text notes. Not real: predicting prices. AI over your trade history is pattern recognition on a sample of a few hundred — useful for hypothesis generation, useless as a forecast. AI trading journal — what's real and what's hype →.
Broker analytics (Zerodha Console and equivalents) are excellent ledgers and deliberately not analytics — they can't be, because they don't know your setups or your planned risk. Use them as the source of truth for fills and charges, then compute on top. Setup guide: Zerodha trade journal guide →, or the direct Zerodha journal import.
14. Comparison
| Capability | Memory / gut | Broker P&L (Console) | Excel / Sheets | TradeDiary |
|---|---|---|---|---|
| Net P&L after all Indian charges | ❌ | partial | ⚠️ DIY formulas | ✅ itemised |
| FIFO matching for ITR | ❌ | ⚠️ tax report only | ⚠️ hand-built | ✅ built-in |
| R-multiple per trade | ❌ | ❌ | ⚠️ if risk logged | ✅ |
| Expectancy (overall + per tag) | ❌ | ❌ | ⚠️ manual | ✅ |
| Max drawdown incl. open positions | ❌ | ❌ | ❌ | ✅ |
| P&L by setup tag / time-of-day | ❌ | ❌ | ⚠️ pivot tables | ✅ |
| F&O legs rolled to strategy level | ❌ | ❌ | ❌ | ✅ |
| Expiry-cycle bucketing | ❌ | ❌ | ❌ | ✅ |
| Behavioural metrics (adherence, tilt) | ❌ | ❌ | ⚠️ manual | ✅ |
| Logs everything (no selection bias) | ❌ | ✅ | ❌ | ✅ |
The honest read: a disciplined spreadsheet covers most of column three, and column three is a long way ahead of column one. The gap that matters is between "gut" and "any consistent measurement at all" — not between a good Sheet and a good app. Start anywhere; just start measuring the same way every week.
15. Case studies
Arjun (Pune, 31, IT, index options 4–5×/week). Roughly breakeven across FY25 on ~₹6L, which felt like a skill problem. It was an arithmetic problem. Rolling his legs into strategy-level positions and recomputing net of charges showed two clean populations: defined-risk spreads at +0.41R expectancy over 118 trades, and naked directional options at −0.37R over 96. Blended, they cancelled — exactly the breakeven he was seeing. He didn't learn a new strategy or change his view on the market; he stopped taking the second population. Next two quarters: +₹1.9L on the same capital, with 44% fewer trades.
Sneha (Kolkata, 29, part-time swing + intraday). Convinced her intraday scalping was carrying the account because those wins felt frequent and immediate. Cost-ratio analysis said otherwise: intraday was 58% of gross P&L consumed by charges, netting −₹18,000 over the year, while 31 swing trades she barely thought about netted +₹1.34L. The screen time was going almost entirely to the losing half. Cutting intraday to two planned setups a week freed the hours and raised net P&L. The finding wasn't a strategy; it was a ratio she'd never computed.
(Both anonymised, details changed; representative of patterns across our user base, not specific individuals.)
How to build your trading analytics stack (6 steps)
- Get a clean, complete ledger — import your tradebook rather than typing it. Selection bias in the input makes every downstream statistic decorative.
- Make P&L net — itemised STT, brokerage, GST, stamp and exchange charges per trade, FIFO-matched for equity. Nothing below this step is trustworthy until this one is right.
- Log planned risk at entry — entry price minus stop, times quantity. Set before the outcome exists, never reconstructed after.
- Tag every trade at entry — one setup tag, and a mistake tag on the ones that deserve it. Ten tags maximum; a scheme you won't maintain produces no data.
- Compute the stack — R-multiple per trade, then win rate, average win/loss R, expectancy and max drawdown. Wait for 30 trades per bucket before believing anything.
- Review weekly, act monthly — 15 minutes each Sunday to sort expectancy by tag; one rule change a month, not five. Kill the worst bucket, size up the best.
🚀 Ready to see your real numbers? Start your free TradeDiary account — auto-import from Zerodha, Kotak, Upstox and Dhan, charges itemised per Indian broker, FIFO-matched and ITR-ready, with expectancy and drawdown computed per setup tag. Free tier covers 50 trades/month — enough to find your first negative-expectancy bucket.
FAQ
What is trading analytics? Trading analytics is the practice of computing statistics on your own closed trades — win rate, R-multiple, expectancy, drawdown — and slicing them by setup, instrument and time so you can identify which parts of your trading make money and which lose it. It differs from market analysis, which studies price; analytics studies you.
What's the difference between a trading journal and trading analytics? The journal is the record — every trade plus the reason you took it. Analytics is the computation on top of that record. You need both: a journal without analytics is a diary you never learn from, and analytics without a journal is just your broker's P&L statement, which knows what you traded but not why.
Which trading metrics actually matter?
Four, in this order: net P&L after all charges, R-multiple, expectancy, and max drawdown. Expectancy — (win% × avg win R) − (loss% × avg loss R) — is the one that tells you whether you have an edge. Win rate is the most tracked and least actionable metric in trading.
How many trades do I need before my analytics means anything? About 30 closed trades per bucket for a rough read, 100+ for confidence. Below 20 the numbers are noise. This is why slicing too finely too early backfires — six one-trade tags tell you nothing, whereas three tags with 40 trades each tell you what to stop doing.
Should I calculate P&L before or after brokerage and taxes? Always after. In India, STT, brokerage, GST, stamp duty and exchange charges frequently consume 30–60% of gross edge for active intraday and options traders. Analytics on gross P&L routinely shows a positive edge where the net result is negative — and that's precisely the situation in which a trader scales up.
What is a good expectancy for an Indian retail trader? Anything reliably positive net of costs is a genuine edge. +0.2R to +0.4R per trade over 100+ trades is a strong, realistic result. Be sceptical of anything above +1R — it usually means risk is being recorded as the actual loss rather than the planned stop, which retrofits every number to the outcome.
How do I analyse F&O and options trades correctly? Roll all legs of a position into one strategy-level net P&L before computing anything, bucket by days-to-expiry and each underlying's native cycle rather than calendar weeks, log entry IV and delta for context, and make sure expired-ITM positions are counted rather than silently dropped. Details in measuring options strategy performance.
Can a spreadsheet do this or do I need software? A spreadsheet is genuinely sufficient under about 20 trades a month with a single broker and no multi-leg F&O — you'll hand-build FIFO matching and a charges formula and maintain them yourself. Beyond that, or with two brokers or any spreads, upkeep exceeds the cost of a tool within weeks. See trade journal vs spreadsheet.
Does trading analytics work for beginners with few trades? Yes, but measure behaviour before performance. With 15 trades your expectancy is meaningless, while your plan-adherence rate and post-loss trade interval are already informative. Behavioural metrics stabilise far faster than performance metrics, which is why beginners should start there.
Can AI do my trading analytics for me? AI is useful for clustering losing trades by factors you didn't think to tag, spotting behavioural drift, and auto-tagging notes. It cannot forecast prices, and any tool claiming to is selling something else. See AI trading journal — what's real.
Will better analytics make me profitable? Not on its own — analytics measures, it doesn't trade. But the losses in SEBI's 9-in-10 figure are overwhelmingly behavioural and cost-related, and both are invisible until counted. Analytics converts an unfixable feeling that something is wrong into a specific, countable thing you can stop doing. That shifts the odds; it doesn't remove the risk.
Risk disclaimer
This article is for educational purposes only and does not constitute investment advice. Trading in equity and derivatives in India carries substantial risk of loss. Past performance is not indicative of future results. As per SEBI's January 2024 study, 9 out of 10 individual F&O traders incurred net losses over FY22. Trade only with capital you can afford to lose, and consult a SEBI-registered investment advisor for personal recommendations.
Author: Pulkit Mangal — Founder, TradeDiary. Equity + F&O trader since 2017. Built TradeDiary after personal losses of ₹14L in FY21 highlighted the absence of a market-aware analytics tool for Indian retail traders.
Last updated: 1 August 2026.