fmp-technical-analysis

# FMP Technical Analysis Skill ## Overview This skill covers the 9 technical indicator tools provided by the `fmp-mcp-server` MCP integration. These tools query Financial Modeling Prep (FMP) for real-time and historical technical indicator data, enabling quantitative analysis across multiple timeframes. ## Available Tools ### Moving Averages #### `fmp_sma` — Simple Moving Average - **Purpose:** Trend direction identification, dynamic support/resistance levels. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Calculates the unweighted arithmetic mean of closing prices over the specified period. Equal weight is given to every data point in the window. - **Typical periods:** 20 (short-term), 50 (medium-term), 200 (long-term). #### `fmp_ema` — Exponential Moving Average - **Purpose:** Trend detection with heavier weighting on recent prices, making it more responsive than SMA. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Applies an exponential decay weighting so that the most recent prices contribute more to the average. - **Typical periods:** 9, 12, 20, 26, 50. #### `fmp_wma` — Weighted Moving Average - **Purpose:** Linear-weighted trend following that prioritizes recent data. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Assigns linearly increasing weights to each data point in the window (the most recent price gets the highest weight). - **Typical periods:** 10, 20, 50. #### `fmp_dema` — Double Exponential Moving Average - **Purpose:** Reduced-lag trend following for faster signal generation. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Combines a single EMA with an EMA-of-EMA to cancel out a portion of the inherent lag. - **Typical periods:** 12, 20, 50. #### `fmp_tema` — Triple Exponential Moving Average - **Purpose:** Minimal-lag trend following for the most responsive moving average signals. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Extends the DEMA concept with a third layer of exponential smoothing, further reducing lag. - **Typical periods:** 12, 20, 50. ### Momentum Indicators #### `fmp_rsi` — Relative Strength Index - **Purpose:** Identify overbought and oversold conditions. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **Interpretation:** Values above **70** suggest overbought conditions; values below **30** suggest oversold conditions. The 50 level acts as a bullish/bearish divider. - **Typical period:** 14. #### `fmp_williams` — Williams %R - **Purpose:** Momentum oscillator measuring the current close relative to the high-low range over a lookback period. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **Interpretation:** Readings above **-20** indicate overbought; readings below **-80** indicate oversold. The scale runs from 0 to -100. - **Typical period:** 14. #### `fmp_adx` — Average Directional Index - **Purpose:** Measure trend strength regardless of direction. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **Interpretation:** Values above **25** indicate a strong trend; values below **20** suggest a weak or range-bound market. ADX does not indicate direction, only strength. - **Typical period:** 14. ### Volatility #### `fmp_stddev` — Standard Deviation - **Purpose:** Quantify price volatility over a given period. - **Key Inputs:** `symbol`, `periodLength`, `timeframe` - **How it works:** Calculates the statistical standard deviation of closing prices, providing a measure of dispersion around the mean. - **Typical periods:** 20, 30. ## Supported Timeframes All 9 tools accept the same set of timeframe values: | Value | Description | |-------|-------------| | `1min` | 1-minute bars | | `5min` | 5-minute bars | | `15min` | 15-minute bars | | `30min` | 30-minute bars | | `1hour` | 1-hour bars | | `4hour` | 4-hour bars | | `1day` | Daily bars | | `1week` | Weekly bars | | `1month` | Monthly bars | Use shorter timeframes (`1min` through `1hour`) for intraday and day-trading analysis. Use longer timeframes (`1day` through `1month`) for swing trading and position analysis. ## Usage Patterns ### 1. Trend Analysis — Golden/Death Cross Detection Combine a short-period SMA with a long-period SMA to detect major trend reversals. ``` fmp_sma(symbol="AAPL", periodLength=50, timeframe="1day") fmp_sma(symbol="AAPL", periodLength=200, timeframe="1day") ``` - **Golden Cross:** The 50-day SMA crosses above the 200-day SMA — bullish signal. - **Death Cross:** The 50-day SMA crosses below the 200-day SMA — bearish signal. Compare the most recent values from both responses. If `sma_50 > sma_200` and the previous period had `sma_50 < sma_200`, a golden cross has just occurred. ### 2. Mean Reversion — RSI + Standard Deviation Use RSI to identify overbought/oversold extremes and standard deviation to add volatility context. ``` fmp_rsi(symbol="MSFT", periodLength=14, timeframe="1day") fmp_stddev(symbol="MSFT", periodLength=20, timeframe="1day") ``` - **High-confidence oversold:** RSI < 30 AND stddev is elevated (price has deviated significantly from mean — reversion is likely). - **High-confidence overbought:** RSI > 70 AND stddev is elevated. - If stddev is low while RSI is extreme, the move may lack the volatility to snap back quickly. ### 3. Momentum Screening — ADX + EMA Filter Filter for stocks in strong trends, then confirm direction with EMA. ``` fmp_adx(symbol="NVDA", periodLength=14, timeframe="1day") fmp_ema(symbol="NVDA", periodLength=20, timeframe="1day") ``` - **Strong uptrend:** ADX > 25 AND current price is above the 20-period EMA. - **Strong downtrend:** ADX > 25 AND current price is below the 20-period EMA. - **Avoid:** ADX < 20 indicates a weak or non-trending market — momentum strategies are less reliable. ### 4. Multi-Timeframe Confirmation Run the same indicator across multiple timeframes to confirm signals and avoid false entries. ``` fmp_rsi(symbol="TSLA", periodLength=14, timeframe="1day") fmp_rsi(symbol="TSLA", periodLength=14, timeframe="1hour") fmp_rsi(symbol="TSLA", periodLength=14, timeframe="15min") ``` - **Strong signal:** All three timeframes agree (e.g., all show oversold below 30). - **Weak signal:** Only the shortest timeframe shows the condition — likely noise. - **General rule:** Higher timeframes carry more weight. A daily oversold reading confirmed by hourly and 15-min is a higher-conviction setup than a 15-min reading alone. ## Best Practices 1. **Always specify `periodLength` and `timeframe` explicitly.** Do not rely on defaults; being explicit ensures reproducible analysis. 2. **Combine indicators from different categories.** A moving average (trend) paired with RSI (momentum) and stddev (volatility) gives a more complete picture than any single indicator alone. 3. **Use multi-timeframe analysis for entry timing.** Identify the trend on a higher timeframe, then time entries on a lower timeframe. 4. **Watch for divergences.** When price makes a new high but RSI or Williams %R does not, a reversal may be forming. 5. **ADX is non-directional.** A rising ADX means the trend is strengthening, but you must use price action or a moving average to determine whether the trend is up or down. 6. **Prefer EMA/DEMA/TEMA over SMA when reduced lag is important** — for example, in fast-moving intraday markets. Use SMA for widely followed levels like the 200-day moving average, where institutional alignment matters.