Práctico 4 - Aprendizaje Supervisado.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c6a3fddb-a42e-4f4f-835a-9ff5e3ea76b1",
   "metadata": {},
   "source": [
    "# Práctico 4: Aprendizaje Supervisado\n",
    "\n",
    "En la industria del trading algorítmico, las etiquetas para predicción no están del todo claras en un comienzo. Además, el hecho de estar trabajando con series de tiempo complica el trabajo. Vamos a dejar un poco de lado este hecho y vamos a ver si podemos generar características útiles con los datos.\n",
    "\n",
    "Comencemos como en todos los prácticos: descargando los datos con los cuales vamos a trabajar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "88034e14-ce54-480d-8cc4-2790f6a5c4b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Attributes</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Close</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Open</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-01-02</th>\n",
       "      <td>610.190002</td>\n",
       "      <td>610.190002</td>\n",
       "      <td>615.840027</td>\n",
       "      <td>576.940002</td>\n",
       "      <td>576.940002</td>\n",
       "      <td>589200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-03</th>\n",
       "      <td>607.799988</td>\n",
       "      <td>607.799988</td>\n",
       "      <td>609.840027</td>\n",
       "      <td>600.059998</td>\n",
       "      <td>602.619995</td>\n",
       "      <td>320900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-06</th>\n",
       "      <td>606.549988</td>\n",
       "      <td>606.549988</td>\n",
       "      <td>613.789978</td>\n",
       "      <td>601.000000</td>\n",
       "      <td>601.039978</td>\n",
       "      <td>209700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-07</th>\n",
       "      <td>625.000000</td>\n",
       "      <td>625.000000</td>\n",
       "      <td>627.969971</td>\n",
       "      <td>604.739990</td>\n",
       "      <td>606.640015</td>\n",
       "      <td>480900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-08</th>\n",
       "      <td>633.440002</td>\n",
       "      <td>633.440002</td>\n",
       "      <td>637.820007</td>\n",
       "      <td>624.549988</td>\n",
       "      <td>625.000000</td>\n",
       "      <td>334600</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Attributes   Adj Close       Close        High         Low        Open  Volume\n",
       "Date                                                                          \n",
       "2020-01-02  610.190002  610.190002  615.840027  576.940002  576.940002  589200\n",
       "2020-01-03  607.799988  607.799988  609.840027  600.059998  602.619995  320900\n",
       "2020-01-06  606.549988  606.549988  613.789978  601.000000  601.039978  209700\n",
       "2020-01-07  625.000000  625.000000  627.969971  604.739990  606.640015  480900\n",
       "2020-01-08  633.440002  633.440002  637.820007  624.549988  625.000000  334600"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pandas_datareader import data\n",
    "\n",
    "tickers = ['MELI']\n",
    "\n",
    "# Definimos el rango de fechas deseado\n",
    "start_date = '2020-01-01'\n",
    "end_date = '2020-12-31'\n",
    "\n",
    "# Obtenemos un DataFrame\n",
    "df = data.DataReader(\n",
    "    tickers, \n",
    "    start=start_date, \n",
    "    end=end_date, \n",
    "    data_source='yahoo',\n",
    ").droplevel(\"Symbols\", axis=1)\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8d3cd42-b553-4068-b0a2-20e03f72d15f",
   "metadata": {},
   "source": [
    "### Hagamos Análisis Técnico!\n",
    "\n",
    "El *análisis técnico* es el arte (algunas personas le dicen ciencia) de predecir los movimientos futuros de los activos financieros en base a su comportamiento en el pasado. Como con la predicción del clima, puede no ser preciso, pero quizás pueda dar algunas ideas de cómo se pueden mover ciertos activos y nos permita tomar decisiones de inversión a partir de esto.\n",
    "\n",
    "Vamos a tener ciertas suposiciones claves para hacer con respecto a esto:\n",
    "- **Alta liquidez**: Se asume que las acciones con las que se trabaja tienen un alto volumen de negociación. Si una acción se negocia mucho, es mucho más difícil manipular su precio, por lo que seguiría un curso \"de manual\", si es que se encuentra ese manual...\n",
    "- **No hay cambios artificiales de precios**: Las acciones pagan dividendos (como el cupón de un bono) o, en algunos casos, pueden ser divididas (mediante una acción que se denomina \"split\"). Estas acciones pueden cambiar de manera artificial el precio de una acción. Para poder tener esto en cuenta, es necesario tener el precio de una acción \"ajustado\" por sus movimientos artificiales. Por suerte, nuestra fuente de datos ya soluciona esto.\n",
    "- **No hay \"cisnes negros\"**: El análisis técnico no puede tener en cuenta noticias externas que puedan afectar al desempeño de una acción de manera sorpresiva, como por ejemplo la muerte de su CEO o un ataque terrorista en el que se viera afectada (o tal vez una pandemia).\n",
    "\n",
    "En resumen, el análisis técnico utiliza la información capturada en el precio de un activo para interpretar qué es lo que el mercado dice, para formar una visión del futuro.\n",
    "\n",
    "Claramente, esto puede sonar a chamuyo, pero puede ser algo de qué agarrarse cuando no se tiene mucha más información con la cual trabajar, ya que ha mostrado ser de utilidad en mercados muy grandes, como en EE.UU.\n",
    "\n",
    "Otra cosa a tener en cuenta: Los datos históricos de mercado generan **series de tiempo**, que son un fenómeno un poco lejano a los datos estáticos. En este práctico vamos a probar las herramientas aprendidas en la teoría y ver si pueden tener incidencia a la hora de trabajar con estas series de tiempo.\n",
    "\n",
    "Vamos a hacer todo esto primero con las acciones de Mercado Libre, para tener una idea de cómo se ve, aplicando la librería de pandas `ta` (https://github.com/bukosabino/ta)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3f702f9a-82be-4e43-b8e6-d33f38ff3698",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting ta\n",
      "  Downloading ta-0.7.0.tar.gz (25 kB)\n",
      "Requirement already satisfied: numpy in /home/lbiedma/miniconda3/envs/diplofce/lib/python3.9/site-packages (from ta) (1.20.1)\n",
      "Requirement already satisfied: pandas in /home/lbiedma/miniconda3/envs/diplofce/lib/python3.9/site-packages (from ta) (1.3.0)\n",
      "Requirement already satisfied: python-dateutil>=2.7.3 in /home/lbiedma/miniconda3/envs/diplofce/lib/python3.9/site-packages (from pandas->ta) (2.8.1)\n",
      "Requirement already satisfied: pytz>=2017.3 in /home/lbiedma/miniconda3/envs/diplofce/lib/python3.9/site-packages (from pandas->ta) (2021.1)\n",
      "Requirement already satisfied: six>=1.5 in /home/lbiedma/miniconda3/envs/diplofce/lib/python3.9/site-packages (from python-dateutil>=2.7.3->pandas->ta) (1.15.0)\n",
      "Building wheels for collected packages: ta\n",
      "  Building wheel for ta (setup.py) ... \u001b[?25ldone\n",
      "\u001b[?25h  Created wheel for ta: filename=ta-0.7.0-py3-none-any.whl size=28718 sha256=467dfd07ce12925c13f1e24bc5945e5c6b594cbb658ad171d2ba668d2c9bb6b7\n",
      "  Stored in directory: /home/lbiedma/.cache/pip/wheels/ba/ce/af/165aa165b3083f07486ae2948c961da8ca903a12bcc0f907f6\n",
      "Successfully built ta\n",
      "Installing collected packages: ta\n",
      "Successfully installed ta-0.7.0\n"
     ]
    }
   ],
   "source": [
    "!pip install ta"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9bebfc0-ba93-4ae8-91e5-fb99fcd61489",
   "metadata": {},
   "source": [
    "# Indicadores Técnicos #\n",
    "\n",
    "La librería tiene varios de los más usados, veamos:\n",
    "\n",
    "- Medias Móviles: Se hace un promedio de los precios anteriores al día que se está tomando. Existen dos de las medias más comunes. La media *simple* (SMA) hace un promedio simple de los precios pasados, mientras que la media *exponencial* (EMA) multiplica cada elemento por un peso que va creciendo de manera exponencial hasta el presente. Aquí el hiperparámetro importante es la cantidad de períodos hacia atrás que se usan en el cálculo. Las medias móviles son indicadores de *tendencia* del precio de un activo (https://www.investopedia.com/terms/m/movingaverage.asp).\n",
    "- Bandas de Bollinger (BB): Indicador de volatilidad, Una vez se tiene la media móvil, se dibuja además un valor de dos desviaciones estándar hacia arriba y hacia abajo del precio del activo (https://en.wikipedia.org/wiki/Bollinger_Bands).\n",
    "- Línea de Acumulación/Distribución (ADL/I): Indicador que utiliza el volumen de venta, junto con los precios del activo. Su fórmula de cálculo se encuentra en https://www.investopedia.com/terms/a/accumulationdistribution.asp. Ayuda a ver cuán fuerte es una tendencia, ya que si un precio sube y este indicador no lo hace, quiere decir que no hay suficiente ímpetu como para mantener la suba y puede volver a bajar.\n",
    "\n",
    "Existen mucho más ejemplos, los cuales están en la documentación y la librería, probemos uno."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "443a6885-e789-43f0-bb5b-d831488ef8b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import ta\n",
    "df[\"ubol\"] = ta.volatility.bollinger_hband(df[\"Adj Close\"])\n",
    "df[\"lbol\"] = ta.volatility.bollinger_lband(df[\"Adj Close\"])\n",
    "df[[\"Adj Close\", \"ubol\", \"lbol\"]].plot(figsize=(14,9))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5afb6e17-9fad-4c40-a8dc-90dfacd14774",
   "metadata": {},
   "source": [
    "# Backtesting de Estrategias\n",
    "\n",
    "Lo importante aquí es aprender cómo actuar a partir de cada señal que se recibe. Para eso comenzaremos a hacer *backtesting*, es decir, vamos a implementar una estrategia de inversión y la vamos a testear en el pasado, para ver si tiene sentido aplicarla.\n",
    "\n",
    "Por ejemplo, la \"teoría\" de las bandas de Bollinger dice que las señales de compra se emiten cuando el precio está por debajo de la banda inferior y habría que vender el activo cuando el precio se va por encima de la banda superior. Es posible hacer de forma automática la señal de las bandas con la librería `ta`, el ejemplo está en su github. Para juntar información de más de una media móvil se puede usar el indicador MACD.\n",
    "\n",
    "Tenemos que etiquetar nuestros datos si queremos aplicar un poco de aprendizaje supervisado! Para eso, vamos a crear etiquetas simples, para saber si al comprar la acción se ganó dinero. Vamos a tomar el precio de la acción en una fecha y compararlo con el precio en una fecha hacia adelante. Podemos usar una semana, o un mes, o quizás hasta un año. Por ahora, hagamos estrategias diarias y etiquetemos con 1 si, al día siguiente, el precio subió o -1 si el precio bajó."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8c662256-5872-45f3-8e42-5c924bc280d1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "720d1318-f404-47ea-a87c-44523b02cf94",
   "metadata": {},
   "source": [
    "Una vez las etiquetas estén listas, comencemos a juntar señales, que van a ser nuestros features! Utilizar algunas de la lista que figura en el Github de `ta`, son muchas, así que tomar algunas que parezcan interesantes o que pueden agregar variabilidad a los resultados (cómo podríamos ver eso?):\n",
    "\n",
    "#### Volume\n",
    "\n",
    "* Accumulation/Distribution Index (ADI)\n",
    "* On-Balance Volume (OBV)\n",
    "* Chaikin Money Flow (CMF)\n",
    "* Force Index (FI)\n",
    "* Ease of Movement (EoM, EMV)\n",
    "* Volume-price Trend (VPT)\n",
    "* Negative Volume Index (NVI)\n",
    "\n",
    "#### Volatility\n",
    "\n",
    "* Average True Range (ATR)\n",
    "* Bollinger Bands (BB)\n",
    "* Keltner Channel (KC)\n",
    "* Donchian Channel (DC)\n",
    "\n",
    "#### Trend\n",
    "\n",
    "* Moving Average Convergence Divergence (MACD)\n",
    "* Average Directional Movement Index (ADX)\n",
    "* Vortex Indicator (VI)\n",
    "* Trix (TRIX)\n",
    "* Mass Index (MI)\n",
    "* Commodity Channel Index (CCI)\n",
    "* Detrended Price Oscillator (DPO)\n",
    "* KST Oscillator (KST)\n",
    "* Ichimoku Kinkō Hyō (Ichimoku)\n",
    "\n",
    "#### Momentum\n",
    "\n",
    "* Money Flow Index (MFI)\n",
    "* Relative Strength Index (RSI)\n",
    "* True strength index (TSI)\n",
    "* Ultimate Oscillator (UO)\n",
    "* Stochastic Oscillator (SR)\n",
    "* Williams %R (WR)\n",
    "* Awesome Oscillator (AO)\n",
    "\n",
    "#### Others\n",
    "\n",
    "* Daily Return (DR)\n",
    "* Daily Log Return (DLR)\n",
    "* Cumulative Return (CR)\n",
    "\n",
    "Vamos a dejarlas en el mismo DataFrame, como hicimos con las bandas de Bollinger. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1b42617-7cf6-4617-9c79-2cd0cb31c1d6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "e080ae36-5add-419c-8a4a-8adbda6e7fc0",
   "metadata": {},
   "source": [
    "## ¿Cuáles parecen ser las mejores señales para inferir 'algo'?\n",
    "\n",
    "Es probable que las mismas cambien a medida que pase el tiempo. Algo *muy importante* al trabajar con datos financieros es separar de forma correcta los datos de training y test, ya que siempre queremos hacer inferencias del futuro en base a datos del pasado (https://www.datapred.com/blog/the-basics-of-backtesting).\n",
    "\n",
    "Vamos a dividir nuestros datos de forma que podamos entrenar una predicción cada trimestre del año 2020. Para eso vamos a poder usar los datos desde comienzo del 2020.\n",
    "\n",
    "Hagamos clasificación binaria, ya que tenemos nuestras etiquetas +1 y -1. K Nearest Neighbors parece interesante para hacer esto, pero también se puede probar algún otro clasificador.\n",
    "\n",
    "Reportar para cada trimestre:\n",
    "\n",
    "- Accuracy\n",
    "- Precision\n",
    "- Recall\n",
    "- F1\n",
    "- matriz de confusión"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94b1c115-da90-4fae-ad8b-034a501e8458",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "559b17b3-a3db-4d43-a62d-55290c80094c",
   "metadata": {},
   "source": [
    "## Finalmente, probar la estrategia\n",
    "\n",
    "Si tomamos la decisión de comprar la acción cuando la etiqueta es +1 y vender (quedarnos con el dinero) cuando la etiqueta es -1 al comienzo de cada día desde el 01/04/2020, cómo nos habrá ido? Supongamos que comenzamos con \\$10.000 y tomemos la diferencia porcentual entre el precio de Cierre del primer día de cada mes para ver si ganamos dinero. Con esto completaríamos un flow de backtesting y vamos a poder evaluar si esta estrategia funciona con la acción MELI.\n",
    "\n",
    "Si hay tiempo, se puede usar esto para otras acciones."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6eb77132-ced3-4af7-977f-734e0178f83a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}