Statistical Edge in Futures Trading: How to Define, Measure, and Defend What You Think You Have

Overview #

Most futures traders who've been at it for a year think they have edge. Almost none of them have measured it. That's not an insult — it's just the uncomfortable gap between "my approach makes sense" and "I can prove, statistically, that my expected value per trade is positive."

Edge isn't a feeling. It's not a streak. It's not the fact that you made money last month. Edge is a measurable, positive expected return per unit of risk over a statistically sufficient sample, after accounting for every real cost of trading. That definition is worth sitting with because everything else in this article flows from it.

If you're a futures trader who suspects you have edge but has never rigorously tested that belief — this article is for you. By the end, you'll know exactly how to calculate your edge, how much data you need to believe the calculation, which futures markets offer structural edges that retail traders can access, and when your edge has decayed to the point where trading it is gambling.

What Statistical Edge Actually Is #

Start with what edge is NOT.

Edge is not "my setup works most of the time." A coin flipped 10 times coming up heads 7 times doesn't prove the coin is biased. Edge is not your most recent month of profitability. Edge is not your conviction about why a market should move. Edge is not even a high win rate — a 70% win rate on trades with a 0.5:1 reward-to-risk ratio loses money.

Edge is a property of a trading approach that produces a positive expected value per unit of risk when measured across enough trades to separate signal from noise.

More precisely: edge = E[P&L per trade] > 0 after commissions, slippage, and exchange fees, when evaluated over a sample large enough to be statistically meaningful.

In futures terms, the cleanest way to express edge is in expected ticks per contract per trade, net of all costs. For ES (E-mini S&P 500), with a tick value of $12.50, an edge of 0.3 ticks/trade means you expect to collect $3.75 per contract per trade on average — before you size up. For NQ, where a tick is worth $5, 0.3 ticks = $1.50. Same mathematical property, radically different dollar impact.

This tick-normalized expression strips away the distraction of dollar amounts and lets you compare edge quality across different instruments and different account sizes.

Why Most Traders Don't Have Measured Edge #

Here's what @rubyslippage put bluntly in a widely-shared NexusFi post: [1]

The gap between hypothesis and measured edge is one of the most important gaps in all of trading. Most traders close their eyes and jump across it.

The Math: Expected Value, Win Rate, and Payoff Ratio #

The foundation of edge is a simple equation:

EV = (Win Rate × Average Win) − (Loss Rate × Average Loss)

Where:

• Win Rate = percentage of trades that close profitably
• Loss Rate = 1 − Win Rate
• Average Win = mean profit per winning trade (in ticks or dollars)
• Average Loss = mean loss per losing trade (in ticks or dollars)

For EV to be positive, you need the right combination of win rate and payoff ratio. These two variables trade off against each other — you can win infrequently with large wins, or win frequently with small ones, but you can't have both a low win rate AND a small average win. Here's what the math actually looks like:

Minimum Viable Combinations (Break-Even)

These are break-even numbers — zero edge. For actual edge, you need to exceed these ratios. A 40% win rate with a 1.5:1 payoff is flat. A 40% win rate with a 1.75:1 payoff is positive edge.

Now apply real costs. For ES, round-trip at retail prices runs approximately $4.50 in commissions plus an average of one tick ($12.50) in slippage per trade = $17 in drag per round trip. If your gross edge is $15/trade, your net edge is negative. You're not making money — the broker and exchange are.

The Commission Reality Check

Take a realistic ES day trading approach: 60% win rate, average win of $100 (8 ticks), average loss of $75 (6 ticks).

• Gross EV = (0.60 × $100) − (0.40 × $75) = $60 − $30 = +$30/trade
• Commission drag: $4.50/round trip
• Slippage drag: $12.50 average (1 tick per trade)
• Net EV = $30 − $4.50 − $12.50 = +$13/trade

That's real edge — thin, but real. Now change the slippage assumption to 2 ticks ($25):

• Net EV = $30 − $4.50 − $25 = +$0.50/trade

Nearly gone. This is why slippage modeling is not a detail — it's a core input to whether you have edge at all.

At prop firm pricing ($0.25/round trip), the math changes dramatically:

• Net EV = $30 − $0.25 − $12.50 = +$17.25/trade

Same strategy, 33% more net edge, purely from execution cost differences.

The High Win Rate Trap #

@SMCJB, one of NexusFi's most analytically sharp contributors, has written extensively about position management mathematics. His analysis highlights a persistent cognitive trap: traders fixate on win rate because winning feels good, even when a high win rate masks poor edge. [2]

A 75% win rate sounds impressive. With an average win of 4 ticks and an average loss of 12 ticks on ES: EV = (0.75 × $50) − (0.25 × $150) = $37.50 − $37.50 = zero. That 75% win rate trader is working hard for nothing.

Loss aversion drives this pattern — traders cut winners short (keeping win rate high) and hold losers too long (keeping loss rate low but making losses bigger). The result: a win rate that feels validating but edge that's underwater.

The Luck vs. Skill Problem: Statistical Significance #

This is the most uncomfortable section of the article, so take a breath.

Suppose you've traded for 4 months, taken 60 trades, and your win rate is 58%. Is that edge? Almost certainly not — and here's the math.

How Many Trades Do You Need? #

For win rate analysis, the standard tool is the binomial test. Under the null hypothesis that you have no edge (50/50 win rate), what's the probability of observing a 58% win rate by chance?

For a 60-trade sample: the standard error of the win rate estimate is √(0.5 × 0.5 / 60) = 6.45%. A 58% observed rate is only 1.24 standard errors above 50% — well within the zone of normal statistical noise. You'd need to observe this at the 95% confidence level to call it signal, which requires being roughly 2 standard errors above 50% = approximately 63% win rate on that 60-trade sample.

A more useful formula: the sample size needed to detect a win rate edge of size ε (the excess above 50%) at confidence level Z (use 1.96 for 95%):

N ≈ Z² / (4 × ε²)

If you believe your true edge is a 55% win rate (ε = 0.05), detecting it at 95% confidence requires: N = (1.96²) / (4 × 0.05²) = 3.84 / 0.01 = 384 trades

For a 58% win rate (ε = 0.08): N = (1.96²) / (4 × 0.08²) = 3.84 / 0.0256 = 150 trades

The brutal takeaway: most traders evaluate their approaches after 20-50 trades. At that sample size, even a consistent 58% win rate doesn't differentiate from a lucky coin flip. @PandaWarrior, who maintains NexusFi's most detailed trade metrics thread, has documented this pattern across hundreds of trader journal reviews: premature confidence in measured edge is the rule, not the exception. [3]

Sharpe Ratio Significance #

If you're using the Sharpe ratio as your edge metric, the required sample size is even more humbling. The statistical test for Sharpe ratio significance requires data inversely proportional to the Sharpe itself.

For an annualized Sharpe of 0.5 (a decent but modest edge), to confirm it at 95% confidence requires approximately:

Years needed ≈ (2 / Sharpe)²

For SR = 0.5: years = (2/0.5)² = 16 years. That's not practical.

A more realistic formulation from Bailey and López de Prado's work on the "Minimum Backtest Length" (MinBTL): you need the backtest to be at least 2.5 times the strategy's average half-life of decay. For most day trading strategies with sharp regime changes, that means 18-36 months of data minimum before Sharpe estimates are trustworthy.

None of this means you should wait 16 years to trade. It means you should be much more humble about what 6 months of data proves.

Survivorship Bias in Your Own Mental Accounting #

There's another layer to the luck-vs-skill problem that no formula captures: traders selectively remember winning periods and reframe losing periods. @tigertrader's observation cuts through the noise: [4]

This is survivorship bias at the individual level. You remember the setups that worked. You forget the identical setups that failed. You construct a narrative of why the winners were obvious and the losers were aberrations. The statistical record doesn't care about your narrative.

Measuring Edge: The Metrics That Matter #

Once you have enough trades (100+ is the floor, 300+ is meaningful), these are the metrics that tell you whether you have edge and what it looks like:

Expectancy #

Expectancy = (Win Rate × Avg Win) − (Loss Rate × Avg Loss)

Express it in dollars per contract per trade. For NQ futures:

• 45% win rate, $500 average win, $300 average loss
• Expectancy = (0.45 × $500) − (0.55 × $300) = $225 − $165 = $60/trade/contract
• After $4.75 round-trip (retail): $55.25 net expectancy

Expectancy below $50/trade/contract on ES/NQ at retail pricing is thin enough that execution variance can wipe it out. Expectancy above $150/trade/contract is strong enough to withstand realistic adverse conditions.

Profit Factor #

Profit Factor = Total Gross Profit / Total Gross Loss

Interpreted as follows:

• PF = 1.0: break-even
• PF 1.1--1.3: marginal edge, typical of many retail discretionary approaches
• PF 1.3--1.7: solid edge — survivable against realistic slippage assumptions
• PF 1.7--2.5: strong edge — approach has structural advantages
• PF > 2.5: very strong — verify you're not overfitting

@Fat Tails, one of NexusFi's most rigorous quantitative contributors, frames it this way: a system with 2,000 trades and a net expectancy of $5/trade after costs is FAR more reliable than a system with 200 trades and $50 expectancy. The statistical confidence in the former is orders of magnitude higher. [5]

Sharpe Ratio (Futures-Adjusted) #

Sharpe = (Mean Return − Risk-Free Rate) / Standard Deviation of Returns

For daily trading PnL on a futures account:

• Compute daily returns normalized to account value
• Annualize by multiplying mean daily return by 252 and daily StdDev by √252
• Subtract the risk-free rate (currently ~5.25% annualized)

Target thresholds for futures:

• SR < 0.5: weak edge, high probability of random variation
• SR 0.5--1.0: acceptable, confirm with larger sample
• SR 1.0--2.0: strong, consistent edge
• SR > 2.0: very strong — verify for overfitting (check for overfitting)

Sortino Ratio #

The Sortino ratio replaces total standard deviation with downside deviation — only volatility below the target return counts. For trend-following futures strategies (where upside volatility is desirable), Sortino gives a more honest picture than Sharpe.

Sortino = (Mean Return − Target) / Downside Deviation

A strategy with occasional large wins but small, controlled losses will show Sortino >> Sharpe. A strategy with symmetric volatility (equally large wins and losses) will show Sortino ≈ Sharpe.

Calmar Ratio #

Calmar = Annualized Return / Maximum Drawdown

For futures accounts where drawdowns are severe, Calmar grounds you in the actual cost of pursuing edge. A Calmar of 1.0 means you earn back your worst drawdown in a year — breakeven from a risk-adjusted perspective. Target Calmar > 2.0 for strong edge; anything below 0.5 suggests the edge isn't worth the psychological and financial cost.

K-Ratio and Equity Curve Diagnostics #

The K-Ratio measures the consistency of the equity curve slope over time, penalizing erratic equity growth. A high K-Ratio means the edge is expressed consistently across different market periods — not just in one regime.

For equity curve analysis more broadly:

• Linear regression of cumulative PnL vs. time: the R² tells you how consistently edge is being expressed (R² > 0.90 = excellent consistency, R² < 0.70 = regime-dependent)
• Drawdown duration is often more diagnostic than magnitude: a 15% drawdown that lasts 2 weeks reflects better edge resilience than a 10% drawdown lasting 6 months
• Recovery factor = net profit / max drawdown. Anything below 1.5 means the edge isn't generating enough return to justify the worst historical pain

In-Sample vs. Out-of-Sample: Where Edges Die #

This is where most "edges" are killed — not by the market, but by honest math.

The fundamental problem: if you develop and evaluate a strategy on the same historical data, you will find edge even in random noise. This is not a bug in your analysis — it's a mathematical guarantee. With enough parameters, any model can be fit to past data. The question is whether that fit reflects genuine market structure or accidental pattern recognition.

The Walk-Forward Test #

Walk-forward validation is the minimum credible test for futures edge:

1. Take your full historical dataset
2. Use the first 70% as "in-sample" development data
3. Develop and improve your approach on that data only
4. Evaluate, unchanged, on the remaining 30% ("out-of-sample")
5. Compare performance: if out-of-sample degrades by more than 50% of in-sample performance, the edge is likely overfitted

The rolling version is more rigorous: slide a fixed-length window forward through history, re-improve at each step, and evaluate on the next period. The distribution of out-of-sample returns across windows tells you whether the edge is strong.

The 3x Degradation Rule #

A practical red flag: if out-of-sample performance degrades by more than 3x relative to in-sample, the approach is almost certainly overfitted. Example:

• In-sample: 65% win rate, profit factor 1.8
• Out-of-sample: 54% win rate, profit factor 1.2

The win rate degraded by 17% and profit factor dropped to 67% of in-sample. This is borderline — the edge may still be real but substantially overstated. If out-of-sample profit factor dropped below 1.0 (net negative), the edge was entirely statistical artifact.

Why Paper Trading Is Not Out-of-Sample #

Paper trading after the strategy is developed does NOT replicate true out-of-sample testing. When you know the setup you're watching for, confirmation bias operates — you paper trade more selectively than you would mechanically. You remember the papers where the setup worked and attribute failures to "execution differences."

True out-of-sample requires data that existed before you looked at it and can't be retroactively adjusted. Pre-COVID data evaluated against post-COVID markets, for example. Or the first 2021-2022 stretch tested against 2023-2024 walk-forward validation.

@matthew28 captures the experience common to many traders with years of effort: the realization that what worked conceptually didn't survive rigorous out-of-sample testing is one of the most common paths to eventual profitability — it forces structural discipline. [10] The cold reality: by rigorous walk-forward standards, most retail approaches that look profitable in backtesting show marginal or negative out-of-sample performance. That's not a failure unique to retail traders — it's a statistical law that affects professional systematic funds too. The difference is that professionals run formal out-of-sample processes before committing capital.

Futures-Specific Edge Sources #

Generic "trading edge" articles talk about momentum and mean reversion without specifying where the structural advantage actually comes from. Futures markets have specific, well-documented sources of structural edge that don't exist in the same form in equities. These are worth knowing in detail.

Roll Yield and Carry #

Futures contracts expire. When the front-month contract is more expensive than the back-month (backwardation), rolling the position forward produces a profit — you sell the more expensive near-term contract and buy the cheaper later-dated one. This is "positive roll yield."

The most famous systematic exploitation of this is short VIX futures (VX). VIX futures almost always trade in contango — the back month is consistently more expensive than spot VIX — and rolling short VX has been a historically consistent source of edge, extracting a volatility risk premium that exists because institutional hedgers willingly pay above fair value for tail protection.

Commodity futures exhibit similar structural patterns:

• Crude oil (CL): frequently in contango when inventories are high, creating roll cost for long holders and roll benefit for systematic short-rollers
• Natural gas (NG): persistent seasonal backwardation in winter months (near-term scarcity) with contango in spring/summer (oversupply); calendar spread strategies built on this seasonal structure have decades of statistical support
• Grain futures (ZC, ZS, ZW): crop year patterns create predictable term structure shapes around planting and harvest cycles

These are not obscure academic findings — they're acknowledged by CME researchers and have been confirmed by independent academic studies across 30+ years of futures data.

Index Rebalancing Effects #

Quarterly reconstitution of the S&P 500 and Nasdaq 100 creates predictable, mechanical order flow in ES and NQ futures. When large-cap stocks are added to or removed from the index, passive funds must buy or sell to maintain tracking. The timing and magnitude of these flows is partially predictable from index methodology.

Research consistently shows excess returns for strategies that position ahead of rebalancing events — not from information asymmetry but from providing liquidity to the mechanical, price-insensitive order flow from index funds. This is a real, exploitable edge, though it requires careful execution and has narrowed as more capital pursues it.

Physical Delivery Pressure #

In deliverable commodity futures (CL, GC, ZC, ZS, NG, ZW), commercial participants must actively manage their physical delivery obligations. As contracts approach first notice day, commercial hedgers who cannot take delivery must roll or close positions — often in size, often at disadvantageous prices.

This creates predictable price pressure patterns:

• Longs who can't take physical delivery liquidate as expiration approaches (price pressure toward fair value basis)
• Convergence of futures price to spot price in the delivery window creates a known, directional process
• Calendar spread positions that capture this convergence have documented edge with clear expiration timing

Seasonal Patterns #

Futures markets have genuine seasonal edge sources driven by physical supply-demand cycles. These are not "sell in May" stock market folklore — they're rooted in real-world supply cycles with decades of statistical confirmation.

Key examples with historical win rates (based on 20+ years of data):

• Natural gas (NG): Long NG in late October/early November ahead of winter withdrawal season — historically positive approximately 70% of years since 2000, with the losses concentrated in anomalously warm winters (El Niño years)
• Corn futures (ZC): Weather premium in June-July (uncertainty about growing conditions) with typical resolution lower by August as crop condition reports clarify supply
• Equity futures (ES/NQ): Pre-holiday bullish bias (last 3 trading days before major holidays) shows statistically significant positive drift; quarterly OpEx weeks show elevated volatility with specific intraday timing patterns

These seasonal patterns survive rigorous walk-forward testing because they're anchored in structural causes, not in data-mining artifacts.

The Liquidity Premium #

Less-liquid futures contracts offer a real edge for traders willing to accept wider bid-ask spreads and lower daily volume:

• ZB (30-year T-bond) vs. ZN (10-year T-note): ZB is substantially less liquid than ZN; the same market-structure setups tend to produce larger per-contract moves in ZB, creating higher gross edge per tick of risk
• RTY (Russell 2000 mini) vs. ES: RTY's relative illiquidity means that when large institutional flows hit, price impact is more pronounced and mean-reversion setups are sharper
• Micro futures (MES, MNQ, MCL): the micro contracts often show slightly less efficient pricing than their full-size counterparts during off-hours, creating brief scalping edge for retail traders that wouldn't exist in the more heavily-arbitraged standard contracts

The liquidity premium isn't free — you bear execution risk and wider realized spreads. But for appropriately sized traders, it's a structural advantage over purely algorithmic competition.

Edge Decay and Regime Dependency #

Here's the truth that every new edge discoverer needs to hear: edges don't last forever.

Market structure evolves. When an edge becomes known, capital flows in to exploit it and compresses the return until it disappears or becomes too small to net positive. This is edge decay, and it operates on different timescales for different types of edge:

• High-frequency microstructure edges: days to weeks. Once spotted by HFT firms, gone.
• Statistical pattern edges (momentum, mean reversion): months to a few years. Alpha gets diluted as strategies proliferate.
• Structural/fundamental edges (carry, seasonality): years to decades, sometimes permanently diminished but rarely eliminated because the underlying cause is persistent (VX contango is driven by structural demand for options, not just trading activity).

Detecting Edge Decay in Real Time #

The practical monitoring approach is rolling window analysis:

1. Track your win rate and profit factor on a rolling 50-trade basis
2. Set warning thresholds based on in-sample performance: if the rolling metric drops more than 30% below your historical mean for 3 consecutive windows, treat it as a regime change signal
3. Reduce size immediately, do not add to losing approaches, switch to paper trading while you reassess

@artemiso, who brings quantitative rigor uncommon on retail forums, has noted that most retail traders confuse a deteriorating edge with "a rough patch" and continue full sizing right through structural decay — this is one of the most costly cognitive errors in active trading. [6]

Regime Dependency: The Hidden Edge Killer #

Many strategies that appear to have edge are actually regime bets — they work in specific market conditions (trending, range-bound, high-volatility) and fail in others. This isn't naturally wrong if you can identify the regime and only trade during favorable conditions. It IS wrong to mistake regime-specific performance for unconditional edge.

A day trading approach that works beautifully during ES trend days (which occur approximately 20-25% of trading days based on Market Profile classification) but bleeds slowly on balance days is not a broadly edged approach. It's a trend-day-only strategy. Evaluated over all sessions, its expectancy might be flat or negative.

The honest assessment: identify which market conditions your approach requires, measure your edge conditional on those conditions, and build explicit regime detection into your process. Trade it when conditions are favorable; sit on your hands when they're not.

Execution Reality: From Simulated Edge to Live Edge #

The single most reliable source of performance degradation from backtest to live trading is execution modeling. Most backtests assume fills at the signal price or at the next bar's open — neither of which represents actual CME Globex execution.

Slippage: The Real Number #

For active ES traders executing at market:

• During RTH liquid hours (8:30 AM CT to 3:00 PM CT): 1 tick ($12.50) average slippage per round trip is a realistic conservative estimate
• During pre-market and post-market: 2-3 ticks per round trip is common
• During economic releases (NFP, FOMC): 3-5 ticks is realistic, with 10+ ticks possible on initial spike

Limit order strategies have different slippage but a different cost: fill probability. If your edge assumes getting filled on limit orders at specific price levels, you need to model the probability that you don't get filled when price blows through your level without resting. A signal that calls for a long at 6920.00 on ES that instead needs a 6919.75 fill because 6920.00 printed for only one contract before moving doesn't get filled. That missed trade represents negative performance that doesn't appear in simulations.

The practical solution: run scenarios. What does performance look like with 0.5 tick slippage? 1 tick? 2 ticks? If the strategy only works with 0-tick slippage, it doesn't work live.

Degradation Example #

Take a medium-frequency ES scalping approach:

• Gross: 300 trades, 62% win rate, average win 6 ticks, average loss 4 ticks
• Gross EV = (0.62 × $75) − (0.38 × $50) = $46.50 − $19 = $27.50/trade
• With 1-tick slippage + $4.50 commission: $27.50 − $12.50 − $4.50 = $10.50/trade
• With 2-tick slippage: $27.50 − $25 − $4.50 = −$2/trade (NEGATIVE)

The strategy lives or dies on slippage modeling. This isn't a catastrophic failure — it's a normal finding that tells you the gross edge needs to be higher, or execution precision needs to be better.

The Equity Curve as Edge Diagnostic #

Once you have enough live data, the equity curve itself becomes a diagnostic tool.

Signs of genuine, consistent edge:

• Approximately linear upward slope (consistent value extraction over time)
• Drawdowns that are relatively brief (weeks, not months) and recover quickly
• R² of equity vs. time regression > 0.85
• Recovery factor > 2.0

Red flags for regime-dependent or decaying edge:

• Long flat periods (months underwater) — suggests the edge only works in specific conditions that aren't present
• Sawtooth equity curve (sharp runups followed by sharp drops) — suggests correlation to a specific market regime rather than consistent extraction
• Deepening drawdowns with each new low — suggests edge is deteriorating
• Equity volatility increasing over time — suggests something has changed in execution or market conditions

The Stopping Rule #

@jamiej83, who contributed extensively to NexusFi's risk management threads, makes a point that applies directly here: trading costs compound against you relentlessly, and the only way to maintain net-positive edge is systematic monitoring. [9] When should you stop trading a strategy based on equity curve signals? The quantitative answer depends on the strategy's parameters, but a practical framework:

• If the rolling 50-trade profit factor drops below 1.0 for 3 consecutive windows: STOP. Reduce to zero size immediately.
• If the drawdown exceeds 2x the historical maximum drawdown: STOP and reassess.
• If the drawdown duration (time underwater) exceeds 2x the historical maximum duration: WARNING. Reduce size 50%, continue monitoring.

Don't trade through these signals hoping for recovery. The statistical probability that the edge has at the core changed is higher than the probability that you're in an unusual but temporary rough patch.

Building Your Edge Framework: Practical Steps #

Start here if you want to apply this systematically:

Step 1: Define the scope — entry, exit, and filter logic specific enough that someone else can replicate every trade without asking questions.

Step 2: Collect data — paper/small size, minimum 100 trades before conclusions, 300+ for real confidence.

Step 3: Run the analysis — expectancy, profit factor, Sharpe, Calmar. Segment by condition: trending vs. ranging, high vs. low VIX, AM vs. PM session.

Step 4: Walk-forward validate — 70% in-sample, 30% out-of-sample. More than 50% degradation means suspect edge.

Step 5: Model execution drag — net expectancy after realistic slippage (1 tick RTH, 2 ticks otherwise) plus your commission tier. Below $15/trade/contract? Improve first, size up later.

Step 6: Set monitoring thresholds — define exactly what rolling metrics cause you to stop. Write it down before you start trading.

Step 6 is the one most traders skip. It's the most important one.

Related Concepts #

Understand First:

• Futures Contract Mechanics — tick size, tick value, contract specs
• Trade Expectancy and Risk/Reward — the foundational calculation

Apply Next:

• Monte Carlo Simulation for Futures Trading
• Walk-Forward Analysis
• Overfitting and Curve-Fitting Detection

Citations #

[1] @rubyslippage, "Dear Ruby," NexusFi, https://nexusfi.com/showthread.php?p=331274

[2] @SMCJB, "Why you should add to winners and never add to losers," NexusFi, https://nexusfi.com/showthread.php?p=489202

[3] @PandaWarrior, "Trading Metrics for journals/record keeping," NexusFi, https://nexusfi.com/showthread.php?p=53940

[4] @tigertrader, "Spoo-nalysis ES e-mini futures S&P 500," NexusFi, https://nexusfi.com/showthread.php?p=539063

[5] @Fat Tails, "Risk of Ruin," NexusFi, https://nexusfi.com/showthread.php?p=210847

[6] @artemiso, "Can Day Trading be profitable for retail?" NexusFi, https://nexusfi.com/showthread.php?p=671485

[7] @tigertrader, "One year later into my trading career," NexusFi, https://nexusfi.com/showthread.php?p=458867

[8] @matthew28, "Traders with 5-10 years of experience but still not profitable," NexusFi, https://nexusfi.com/showthread.php?p=859493

[9] @jamiej83, "Concerning risk per trade sizing," NexusFi, https://nexusfi.com/showthread.php?p=207780

[10] @matthew28, "Traders with 5-10 years of experience but still not profitable," NexusFi, https://nexusfi.com/showthread.php?p=859493