Risk Assessment Suite (NEW)
Risk Assessment Suite
Eight ThinkOrSwim Indicators for Position Sizing, Trade-System Stress Testing, and Edge Visualization
Research Preview | TOS ThinkScript | Free for the Custom Group
What This Suite Is
Most traders treat risk management as an afterthought — a stop loss, a percentage, a number you pick once and forget. The Risk Assessment Suite is built on a different premise: -how you size and stress-test a trade matters as much as the signal that triggered it-.
This is a collection of eight ThinkScript indicators that together let you :
(1) size each trade based on the live probability of being stopped out
(2) translate your historical edge into a Kelly-optimal capital allocation
(3) visualize what your strategy actually looks like under thousands of simulated paths instead of one realized history.
These tools are shared in the spirit of open research. They are not a strategy. They are the mathematical scaffolding that sits underneath any strategy — including the ones already published at MarketFragments.com.
The Eight Indicators
Each indicator stands alone. Used together, they form a complete risk-and-sizing workbench.
1. Monte Carlo Position Sizer (GBM)
The headline tool. On every bar, it fits a Geometric Brownian Motion model to the most recent log-returns of the symbol you're watching, computes the multi-step probability that the next trade entered here will hit its stop loss before its target, and emits a position-size multiplier — full size, three-quarter, half, or quarter — based on that probability.
What's good about it: Most "risk" indicators are static — they pick a percentage and walk away. This one updates continuously with the volatility regime. When markets are quiet and your edge is high, you get full size. When volatility expands and the probability of being stopped out climbs past 65%, the indicator quietly scales you down to a quarter. You don't need to remember to do this. The math does it.
What's risky: GBM is a lognormal model. Real markets fat-tail. When you see a low P(SL) reading, what you're actually seeing is "low under the assumption that returns are lognormal." That assumption breaks during news events and regime breaks. Treat the size multiplier as a regime-aware ceiling, not as a forecast.
2. Kelly Criterion + Monte Carlo + Fixed Fractional Sizer
A single-tool position-sizing workbench. Enter your strategy's win rate and reward/risk ratio, choose between Full Kelly, Half Kelly, Quarter Kelly, or a fixed fractional override, and the indicator simulates one Monte Carlo path of 500 trades through that sizing model — plotting the realized equity curve, max drawdown, max win streak, and max loss streak.
What's good about it: Kelly is famous for being aggressive. Full Kelly is mathematically optimal for terminal wealth, but it produces gut-wrenching drawdowns even on a strong edge. This indicator lets you *see* what those drawdowns look like before you put real money on them. Reload the chart, get a different sample path. Do it ten times. You'll know whether you can stomach Full Kelly or whether you actually need Half.
What's risky: Kelly assumes your win rate and R/R are stable. They aren't. A drift in either parameter — which is normal in live trading — turns "optimal" into "overlevered" fast. The recommended default is Half Kelly, and the indicator flags any input that produces a negative Kelly (no edge).
3. Monte Carlo Trade System Expectancy
The path-dependence visualizer. Given a win rate, R/R ratio, risk percent per trade, and starting capital, it simulates 500 random trades and tracks the realized capital curve, plus your actual win and loss totals, average dollar per trade, and the longest consecutive winning and losing streaks the random sequence happened to produce.
What's good about it: This is the indicator that answers the question "I have an 89% win rate — why am I losing money?" The answer, almost always, is that your max losing streak in the realized sample was longer than your account could survive. This tool exposes that risk numerically. Run it ten times. Note the longest losing streak you see. That's the streak you have to be sized for — not the average.
What's risky: It's one path per chart load. The whole point is to look at many. Don't take a single run as gospel.
4. Monte Carlo Average Parameters
The simplest of the Ehlers-derived stack. Plots a single random cumulative-profit curve next to the deterministic theoretical expectation. The gap between the two is the cost of path dependence — the gap that closes only as the number of trades grows.
What's good about it: Pedagogically clean. If you've never internalized why a positive-expectancy strategy can lose money for stretches that feel infinite, this indicator makes it geometric. The theoretical line goes up smoothly. The realized line wobbles around it. Sometimes for a long time.
What's uncertain: The model uses constant-magnitude wins and losses. Real trades have variable magnitude. See indicator #5 for that.
5. Monte Carlo Randomized Parameters
The same model as #4, but each individual win and loss is given its own random magnitude on top of the win/loss flag. The realized curve becomes noisier and looks more like real per-trade P&L. Useful for showing the additional variance that lives inside each outcome — not just whether you won, but by how much.
6. Monte Carlo Bell Curve (One Symbol)
Builds a normal-distribution bell curve over the per-trade profit distribution from a single-symbol Monte Carlo simulation. Reports mean, standard deviation, and Sharpe ratio of the simulated stream, with a qualitative Sharpe rating ("sub-optimal" through "excellent").
What's good about it: Most traders chase win rate. The bell curve makes the trade-off visible — a high win rate with a fat right tail can have the same Sharpe as a lower win rate with a tighter distribution. Edge isn't a number; it's a shape. This indicator shows you the shape.
What's uncertain: The Sharpe rating thresholds (1.0 / 1.25 / 1.75 / 2.25) are conventional, not strategy-specific. Use them as orientation, not as targets.
7. Monte Carlo Bell Curve (Portfolio — 4 Independent Streams)
The same bell-curve framework as #6, but averaged across four independent Monte Carlo trade streams to mimic the smoothing effect of trading multiple uncorrelated edges in parallel.
What's good about it: Run the One-Symbol version (#6) and the Portfolio version (#7) side-by-side with the same inputs. The portfolio bell curve is visibly narrower and the Sharpe ratio almost always improves. That's diversification, drawn for you.
What's uncertain: "Independent" is the assumption. Real portfolio strategies have correlated drawdowns during regime breaks. The 4-stream model assumes independence, which is the best case. Live results will sit between the One-Symbol and Portfolio curves, closer to the Portfolio side in calm markets and closer to the One-Symbol side during stress events.
8. Fibonacci Risk / Reward Sizer
Auto-detects swing highs and lows over a user-defined length, derives the implied A-B-C Fibonacci structure of the most recent swing, and from that structure computes an entry price, momentum-objective target, trailing stop, retracement percentage, and reward/risk ratio. Equities and futures sizing are both supported — for futures, supply your required margin percent and the indicator returns implied position size in contracts.
What's good about it: Most Fibonacci tools require you to manually anchor the swing. This one finds the swing automatically and updates as new swings form. Useful as a fast structural read on whether the current setup has acceptable reward-to-risk before you commit to it.
What's uncertain: Fibonacci sizing is structural, not statistical. The reward/risk number is derived from the geometry of the most recent swing, not from any historical fill probability. Combine with the Monte Carlo Position Sizer (#1) for a probability-weighted view.
Recommended Setup
There are eight indicators, but you don't need all eight on the chart at once. A typical setup:
On the chart (upper): Fibonacci Risk/Reward Sizer (#8) for structural context. Kelly + MC + Fixed Fractional (#2) if you want a live equity curve overlay.
Below the chart (lower): Monte Carlo Position Sizer (#1) — this is the one you watch every bar. Add Bell Curve (One Symbol, #6) when you're stress-testing a new edge.
Used periodically, not every day: Trade System Expectancy (#3), Average Parameters (#4), Randomized Parameters (#5), Bell Curve Portfolio (#7). These are research tools — load them when you're sizing a new strategy or reviewing an old one, not as a daily overlay.
Inputs Cheat-Sheet
| Indicator | Key Inputs | Notes |
|-----------|-----------|-------|
| MC Position Sizer | num_sims, lookback_params, sl_atr_mult, tp_rr | Defaults tuned for 5-min charts. Larger lookback = more stable P(SL). |
| Kelly + MC + FF | startingCapital, winProb, rewardRisk, sizingModel | sizingModel default is Kelly. Switch to HalfKelly for production. |
| Trade System Expectancy | startingCapital, RiskPercent, WinningPercent, RR, Runs | Reload chart to draw a new sample path. |
| MC Avg Parameters | WinningPercent, profitFactor, Runs | Plotted in the lower panel. |
| MC Randomized Parameters | WinningPercent, profitFactor, Runs | Same as Avg, with magnitude noise. |
| MC Bell Curve (One Symbol) | WinningPercent, profitFactor, Runs | 81-bar bell window at the right edge of the chart. |
| MC Bell Curve (Portfolio) | WinningPercent, profitFactor, Runs | Same window, four-stream averaged. |
| Fibonacci R/R | length, Equity_to_Risk, Risk_Tolerance, Futures, Required_margin | Toggle Futures on for contract sizing. |
Important Disclaimers
All Monte Carlo tools simulate hypothetical trades using user-supplied parameters. They do not analyze live order flow, fills, slippage, or commissions beyond a single flat number per trade. Results are illustrative.
GBM-based probability estimates assume lognormal returns. Real markets exhibit fat tails, especially around scheduled events (FOMC, earnings, NFP). Treat probability outputs as regime-aware *priors*, not forecasts.
Kelly fractions are mathematically optimal for terminal wealth under stable parameters. Live parameter drift makes Full Kelly extremely aggressive in practice. Half Kelly is the conservative-aggressive default used across the industry; Quarter Kelly is appropriate for less-tested edges.
These indicators are research and educational tools. They are not financial advice, not signals to trade, and have not been forward-tested in live markets as a complete system.
Installation
Each indicator is a standalone ThinkScript file. To install:
1. Open ThinkOrSwim → Studies → Edit Studies → New
2. Open the .ts file in any text editor, copy the entire contents
3. Paste into the ThinkScript editor and click OK
4. Apply the study to a chart from the Studies menu
For the Risk Assessment custom group: add all eight studies to a single chart workspace, save the workspace as a template under your Custom Group section, and you have the full suite available with one click.
If you test these tools and have results — or find a bug — we want to hear from you. MarketFragments.com is building toward a community of rigorous, data-driven traders. The Risk Assessment Suite is an invitation to join that work.