Chart (Demo) — Lightweight Charts minimal sample
Minimal Lightweight Charts demonstration. This page renders synthetic OHLC data via TradingView's open-source Lightweight Charts library (~45 KB gzipped, Apache 2.0). NO real prices. NO broker API. NO order routing.
About indicator display
These overlays are mathematical transforms of the synthetic OHLC series, computed in-browser via the technicalindicators library (MIT). They are NOT trading signals, NOT predictions, NOT advice. They visualise what the formulas in the dictionary look like on a price series.
ROC/ROCP/ROCR/ROCR100 (rate_of_change canonical class) and SMA/EMA/WMA (moving_average_kernel) appear as multiple distinct lines but carry the same information modulo an affine map / kernel shape — see /indicator-theory and /honest-confluence for the formal observation.
About these signal markers
These markers (arrows / threshold highlights / collapse overlays) are post-hoc geometric annotations on the synthetic OHLC series. They are NOT trade entries, NOT predictions, NOT advice. They mark events that already happened on the chart — nothing about the future.
The ROC-family collapse demo is the load-bearing point of the dictionary's canonical_class layer: ROC / ROCP / ROCR / ROCR100 plotted on separate price scales look like four distinct indicators, but they are affine transforms of a single ratio R = price / price[t−n]. The shapes are identical; only the y-axis labels differ. N_eff of these four is ≈ 1 — formally, not heuristically. See /indicator-theory canonical-class section.
About Real OHLC mode
Real mode fetches OHLC data from Twelve Data via a Cloudflare Workers proxy with KV cache (5-min TTL) and per-IP rate limit. The data is delayed by roughly one minute and may be stale by up to five minutes due to caching. This is a visualization tool, NOT a trading platform.
NO orders. NO advice. NO buy/sell signals. Indicators and signal markers are mathematical transforms of the displayed series and are NOT predictions. The Twelve Data free tier has a request quota — if you see a fetch error, the page falls back to synthetic data so the chart still renders. (FIEA / 金商法 line-safe.)
What this is NOT
- It is DEMO data (synthetic deterministic random walk seeded from a constant). NOT real market prices.
- It does NOT fetch live prices from any broker, exchange, or data feed in this minimal sample.
- It does NOT support placing orders. There is no broker connection. (FIEA / 金商法 line-safe.)
- It does NOT provide investment advice. No buy/sell signals are displayed.
Technical notes
- Library
tradingview/lightweight-charts (Apache 2.0, open-source)- Gzipped size
- ~45 KB (one of the smallest financial-chart libraries available)
- Performance
- 50,000+ candles render smoothly at 60+ FPS (TradingView official benchmarks). Marker overlays remain comfortable to ~15,000 data points.
- Data source
- Synthetic random walk for this minimal sample. Future versions may add OANDA / Twelve Data / Alpha Vantage via WIC backend proxy (API keys never exposed to browser).
- Architecture
- Display layer (TradingView widget on Home) and analytical layer (this chart + future indicators) are physically separated — both can coexist without conflict. No data redistribution from TV widget.
What should indicator users actually seek? — honest filter discussion
This section articulates the philosophical / mathematical honest filter underlying the very act of displaying indicators on a WIC chart. This is NOT investment advice. It re-frames the well-established academic facts — "indicators are formulas," "markets are close to random walks," "retail traders lose 95%+ over the long term" — from the perspective of the retail user, as educational discussion.
1. Observation: what are indicator users worldwide actually looking for?
Honestly observing the dominant motivation of retail users, it largely collapses to two axes: "searching for always-winning numbers" and "searching for ways to handle sudden moves". This applies to roughly 80-90% of retail users:
| User segment | Indicator usage motivation | "Always-win + spike defense" framing fits? |
|---|---|---|
| Retail FX trader (the majority) | Hunting for "winning combinations" + tuning settings to "guard against spikes" | ✓ Completely |
| MQL5 marketplace buyer | Expecting paid EAs / built-in winning logic | ✓ Yes |
| TradingView Pine Script learner (early stage) | Looking for "their personal high-win-rate setup" | ✓ Yes |
| Side-hustle trader (weekend investor) | "Side income via indicators" | ✓ Yes |
2. The remaining 10-20%: smart users with different motivations
Professional and academic users employ indicators for different purposes. The common thread: they do not believe in an "always-winning indicator":
| User segment | Alternative purpose |
|---|---|
| Professional quant fund (Renaissance, Two Sigma, AQR) | Risk management — ATR for position sizing, VIX for regime detection, strategies built around "marginally positive expectancy × thousands of trades" |
| Macro hedge fund manager (Soros, Druckenmiller etc.) | Confirmation tool — secondary check for discretionary judgment; indicators support, not drive, decisions |
| Factor investor (Fama-French + extensions) | Index construction — momentum / value factors defined via indicator math, used to assemble long-only ETFs |
| Academic researcher (Lo, Mandelbrot, Black-Scholes) | Market microstructure research — the question of "how do indicators model the market" is itself the object of study |
| Technical analyst (Goldman strategist etc.) | Communication tool — chart annotation for clients ("RSI divergence here"); analyst doesn't necessarily believe the signal |
| Risk officer at a bank | Volatility regime monitoring — Bollinger Bands drive margin call thresholds |
3. Core insight: indicators are "measurement," not "prediction" (a category error)
indicator = a number computed from the last N closing prices = measurement tool
always-winning = correctly guessing whether the next bar is up or down = prediction tool
↑ This is the category error
Past measurement ≠ future prediction (Hume's problem of induction).
→ Searching for an "always-winning indicator combination" is therefore a logically impossible search.
"Handling sudden moves" is the same category error: spikes are stochastic / black swan / news shocks — inherently future events that cannot be extracted from past data via mathematical transformation. At best you can do volatility regime detection ("are we in a high-volatility period right now?"), but "will a spike occur in the next 5 minutes?" is not extractable.
4. Why retail traders actually lose — 7 loss factors (spikes are only one)
If the category error explains why "always-winning indicators" cannot exist logically, what then explains the operational reality of retail traders losing money? Many retail users perceive that "spikes are what wipe them out", but honest observation reveals that spike events are just one loss factor among many — and not even the largest one:
| Loss factor | Impact | Details |
|---|---|---|
| (a) Cost structure (spread + slippage + swap) | ★★★★★ | ~1-2 pip spread cost per trade. With 50 trades/month, the strategy edge must be 50-100 pips just to break even. Most indicator-based strategies do not clear this threshold. The broker's spread + slippage on fast moves + overnight swap together silently erode expectancy. |
| (b) Position sizing / excessive leverage | ★★★★★ | Risking 5%+ of capital per trade means 5 consecutive losses → 25% drawdown → very hard to recover (violates Kelly Criterion). Retail FX allows 25× (Japan) / 30× (EU) / 500×+ (offshore); "trade big with little capital" marketing is the structural underbelly. |
| (c) Psychology — cut winners short, hold losers long | ★★★★ | Tversky & Kahneman prospect theory: loss-aversion asymmetry causes profit-taking-too-early + loss-cutting-too-late, which shifts expected value negative even with a 60% win rate. |
| (d) Broker asymmetry (retail FX specific) | ★★★★ | In dealer-model (B-book) retail FX brokers, the user's loss is the broker's profit by direct mechanism. Structurally, losing users are the profitable customers. Even in STP/A-book, spread mark-up + slippage control creates the same incentive. Stop-hunting via signal streaming is theoretically possible. |
| (e) Adverse selection in news/spike events | ★★★ | Stop-loss runs during spikes (NFP / FOMC / central bank speeches). This is the factor most retail traders identify as their primary loss cause — but it is just one of seven, and (a)(b) alone make even 1-2 spike events catastrophic. |
| (f) Strategy decay (alpha disappearance) | ★★★ | Working strategies, once discovered, are arbitraged away (Lo, Adaptive Markets). A backtested 70% win rate drops to ~50% live. Backtests are also prone to survivorship bias. |
| (g) Behavioral / cognitive bias | ★★★ | Confirmation bias (only looking at indicators that agree with your view) / recency bias (over-weighting recent experience) / overtrading (trading out of boredom) / sunk-cost fallacy (unable to cut losses) etc. |
Key observation: many retail traders attribute their losses to "spikes" (factor e), but on honest examination cost structure (a) and position sizing (b) are far larger contributors. Spikes are only the fifth factor; (a)(b)(c)(d) combined explain 90%+ of long-term retail losses. Searching for "an indicator that predicts the next spike" is therefore missing the largest loss factors.
Given these 7 axes, what users should actually seek is not "an indicator that predicts the next spike" but rather tools that systematically measure all 7 axes (a)-(g). The Reframes 1-5 and the 8 metrics that follow are calibrated to this operational reality.
5. Reframe 1: prediction → measurement
| Impossible usage (current) | Possible usage (proper) |
|---|---|
| Predict "is the next bar up or down?" | Measure "what volatility regime are we in?" |
| Discover an "always-winning combination" | Quantify "trend strength over the last N bars" |
| Sense "warning signs of an impending spike" | Recognize "we are in a high-volatility period" |
| Future forecast (impossible) | Present measurement (possible) |
Indicators are measurement instruments. A speedometer doesn't predict "your speed in 5 minutes" — same category.
6. Reframe 2: reading the market → reading yourself
| Attempting to read the market (mostly illusory) | Reading yourself (high-ROI usage) |
|---|---|
| Read chart patterns to call the next move | Monitor your own overtrading frequency with an indicator |
| Decide "this is the bottom" | Quantify your own cut-loss delay pattern |
| Predict impact of news events | Detect your own revenge trades (emotional trades after losses) |
| Hunt for "winning patterns" | Measure your own calibration (actual win rate at 60% confidence) |
| The market (uncontrollable) | Yourself (controllable) |
"Self-analysis using your trade journal + indicators" is the only controllable improvement loop.
7. Reframe 3: winning → not losing
Given that 95% of retail traders lose over the long term:
- "Winning" = being in the 5% = statistically very hard
- "Not losing" = also being in the top 5% = the same statistical fact, inversely framed
- The goal is capital preservation, not "growing the account" (in the short-term trading context)
Concrete metrics:
- Defend maximum drawdown < 10%
- Position size at half of the Kelly Criterion or less
- Mandatory break after 3 consecutive losses
- Monthly trade-count cap (to prevent overtrading)
8. Reframe 4: outcome → process
| Outcome thinking (harmful) | Process thinking (beneficial) |
|---|---|
| "Last month +5%, strategy is working" | "Last month: 25 trades, all per the rules. Outcome is short-term noise." |
| "Last month -3%, strategy is broken" | "Last month: 30 trades, 5 rule violations. What needs fixing is the rule violations." |
| "How much will I earn this month" as a goal | "30 trades this month, all per the rules" as a goal |
| Short-term outcome ≈ noise | Process consistency = the only thing controllable |
Short-term outcomes (win/loss) are nearly random noise. Long-term expected return is determined by process quality (the mathematics of the Sharpe ratio).
9. Reframe 5: certainty → calibration (the most load-bearing)
- "It will definitely go up" / "it will definitely go down" = miscalibrated overconfidence (Dunning-Kruger)
- "60% probability of going up" with an actual win rate of 60% = calibrated
- Probabilistic thinking + expected-value computation = the core skill of professional traders
| Situation | What the user should be seeking |
|---|---|
| When you think "this will go up" | Articulate the full triplet: probability (70%?) + expected value (R:R 1:2?) + bet size (Kelly?) |
| When "indicators align" | Check N_eff (effective independent signal count). If 5 indicators give N_eff = 1, do not raise confidence (this is exactly the honest-confluence-mt4 N_eff demo) |
10. Conclusion: 8 concrete metrics users should seek (mapped to the 7 loss factors)
Putting the 5 reframes together, here are the 8 axes of honest indicator usage (each mapped to the loss factor it addresses):
| # | What to seek | Where existing indicators help | Addresses loss factor |
|---|---|---|---|
| 1 | Risk measure (how much capital is at risk per trade?) | ATR for stop placement / volatility-based position sizing | (b) sizing |
| 2 | Volatility regime (are we in a dangerous period right now?) | BB width / ATR percentile / VIX | (e) spike |
| 3 | Position sizing (how much to bet?) | Kelly Criterion calculator + ATR-based formulas | (b) sizing |
| 4 | Self-auditing (am I following my own rules?) | Trade journal + adherence metrics | (c) psychology + (g) bias |
| 5 | Strategy decay (has the edge disappeared?) | Rolling win rate / Sharpe ratio of the last N trades | (f) decay |
| 6 | Calibration (does confidence match actual win rate?) | Brier score / reliability diagram | (g) bias |
| 7 | Cost awareness (is the spread eating my edge?) | Total cost / total profit ratio | (a) cost |
| 8 | Correlation check (are the signals really independent?) | The honest-confluence N_eff demo | (c) false confidence |
These 8 are honest indicator usage focused on measurement (not prediction), self (not market), and process (not outcome). Note: (d) broker asymmetry cannot be addressed by any indicator — it is a structural choice of broker / dealer-vs-ECN, handled outside the indicator toolkit.
11. Brutally honest aside: "should retail traders trade at all?"
Pure expected-value comparison of long-term returns:
| Option | Expected long-term return | Effort required | Stress |
|---|---|---|---|
| S&P 500 index ETF (buy & hold) | ~7-10% / year (real) | 5 minutes / month | Low |
| NASDAQ index ETF | ~8-12% / year (real) | 5 minutes / month | Low |
| Japan equity index (TOPIX / Nikkei) | ~5-7% / year (real) | 5 minutes / month | Low |
| FX day trading (retail average) | -15% to -50% / year (95% lose) | 50-200 hours / month | High |
| FX swing trading (highly skilled) | -10% to +5% / year | 30-50 hours / month | Medium |
| Equity long-term value (highly skilled) | +5% to +15% / year | 10-20 hours / month | Medium |
On a pure expected-value calculation:
- Seeking "I can win via indicators" = statistically a negative-expected-value game
- Seeking "compound returns via index funds" = positive-expected-value game with minimal effort
So the most rational answer is: "do not seek to trade with indicators (as an income source) in the first place". This is brutally honest, but it is a mathematical truth.
12. Then what value can trading + indicators still have?
If trading is not recommended as an income source, is there other value?
| Motivation | Honest evaluation |
|---|---|
| Income source | Negative expected value — not recommended |
| Side hustle | Same as above |
| Intellectual challenge / mathematical curiosity | ○ Legitimate motivation (same category as chess / poker) |
| Studying market microstructure (academic) | ○ Legitimate (Lo, Mandelbrot, behavioral economics) |
| Self-understanding / psychological exploration | ○ Legitimate (measuring your own biases against the market) |
| Community / hobby | ○ Legitimate (same as a chess club) |
| Entertainment / excitement | △ Same category as a casino — at your own responsibility |
13. The stance of WIC + honest-confluence-mt4
Given this understanding (7 loss factors + category error + 5 reframes), WIC + honest-confluence-mt4 deliberately chooses:
| What WIC does not do | Why (which loss factor) |
|---|---|
| Trade entry / order routing | To not entangle users in (a) cost + (b) sizing + (d) broker asymmetry |
| Claim a "winning indicator combination" | We do not sell what does not exist (rejects the category error) |
| Claim direction prediction | Confidence intervals are too wide — not honest |
| EAs / automated trading | An amplifier of (b) sizing and (c) psychology bypass loss factors |
| IB / affiliate / paid plans | Avoids being embedded in (d) broker incentive structures |
| What WIC does do | Why |
|---|---|
| Visualize indicator formulas (formula → drawing) | Educational + epistemically honest (no magic) |
| ROC family affine collapse demo (mathematical proof of canonical_class) | Destroys the illusion that "5 aligned indicators = 5 signals" (actually N_eff = 1) — directly addresses (c) false confidence and (g) confirmation bias |
| "Does not teach how to win" stance | Refuses to amplify the category error |
14. Core message
References (academic sources cited)
- Fama, E. (1970). "Efficient Capital Markets: A Review of Theory and Empirical Work." Journal of Finance.
- Kahneman, D. & Tversky, A. (1979). "Prospect Theory." Econometrica. / Kahneman, D. (2011). Thinking, Fast and Slow.
- Lo, A. (2008). Adaptive Markets: Financial Evolution at the Speed of Thought.
- Lo, A. & MacKinlay, A.C. (1999). A Non-Random Walk Down Wall Street.
- Mandelbrot, B. (2004). The (Mis)Behavior of Markets.
- Hume, D. (1748). An Enquiry Concerning Human Understanding — problem of induction.
- Kelly, J. (1956). "A New Interpretation of Information Rate." Bell System Technical Journal — the Kelly Criterion.
- Brier, G. (1950). "Verification of Forecasts Expressed in Terms of Probability." — the calibration metric.
- Tversky, A. & Kahneman, D. (1991). "Loss Aversion in Riskless Choice." Quarterly Journal of Economics — theoretical basis for (c) psychology.
- ESMA / FCA / Japan FSA retail FX loss statistics (publicly disclosed).