Technical analysis in R

Introduction

Alvin Tan created interesting Noteboook, which is partly used in this presentation. The Investopedia information source is also used. In addition, a Bookdown of Technical Analysis with R (second edition) from Ko Chiu Yu and ChatGPT were helpful with the coding in R.

Technical analysis is a method to predict price movement in the financial markets. Unlike the usual balance sheet analysis (fundamental investing), technical analysis estimates the value of a given stock using underlying trends of the price movement. The core assumption behind this method is that the fundamentals (information from financial statements) were factored into the price fluctuation, hence, detecting the patterns and signals from the fluctuation should provide sufficient indicator for future performance.

For this project, we are going to explore some of the common technical indicators using the TQQQ Exchange trade fund (ETF) stock. Among leveraged ETFs, ProShares UltraPro QQQ (TQQQ) is one of the largest with assets under management of \(\$18.56\) billion as of July 2022. TQQQ is also one of the more heavily traded leveraged ETFs in the U.S. with an average daily volume of \(\$5.29\) billion (compared with QQQ’s \(\$21\) billion).

TQQQ carries an expense ratio of 0.95%.

#install.packages("quantmod")
#install.packages("data.table")
#install.packages("ggplot2")
#install.packages("tidyquant")
library(quantmod)
library(data.table)
library(ggplot2)
library(tidyquant)  

Data Extraction & Manipulation, Data structure

TQQQ data is extracted using the getSymbols function from Yahoo Finance. Alternatively, the source can be replaced with FRED, MySQL, Google, or others. The data extracted should be in an ‘xts’ object - a format commonly used in time series analysis where observations are recorded in a matrix structure with an index of corresponding time.

Technically, analysis can be done in xts format alone, but for the sake of simplicity, data is converted into a standard dataframe and index column is renamed as “Date’. The end result is shown as below:

getSymbols('KO',src = 'yahoo')       ## change TQQQ ticker to Your ticker symbol

[1] "KO"

structure of the stock data

stock <<- as.data.table(KO)        ###### change TQQQ ticker to Your ticker symbol  #########
names(stock)[1] <- c('Date')

tail(stock)

Interpretation of the HLCO graph

In finance, HLCO graph is a type of financial chart that plots the high, low, close, and open prices for a given period of time. The HLCO chart is similar to a candlestick chart, but it includes the open price in addition to the high, low, and close prices.

The interpretation of an HLCO chart can provide insights into the price action of a security during a given period of time. Here is a brief explanation of how to interpret each component of an HLCO chart:

1. High: The top of the vertical line represents the highest price reached during the given period. This shows the highest price that buyers were willing to pay for the security during the period.

2. Low: The bottom of the vertical line represents the lowest price reached during the given period. This shows the lowest price that sellers were willing to accept for the security during the period.

3. Close: The horizontal line to the right of the vertical line represents the closing price for the period. This shows the final price at which the security traded during the period.

4. Open: The horizontal line to the left of the vertical line represents the opening price for the period. This shows the initial price at which the security traded during the period.

By looking at the patterns of the high, low, close, and open prices in an HLCO chart, traders and investors can gain insight into the overall trend and sentiment of the market for the given security. For example, a long green candlestick with a high close and low open may indicate bullish sentiment, while a long red candlestick with a low close and high open may indicate bearish sentiment.

# feel free to change "HLCO" to "candlestick" or to some other types ("lines", "auto"). 
chartSeries(stock, subset='2020-01-01::2020-09-23', type = "candlesticks", bar.type = "ohlc",theme=chartTheme('white'))

Volume

One of the most common basic indicators traders examine is the trading volume. Trading volume is an indication for the ‘activeness’ of a financial instrument. Depending on the financial instruments, trading volume can be measured either using the number of stocks traded or number of contracts with changed ownerships. To put this in practice, if an increase in volume is observed with a steady increase in price, the instrument can be viewed as steady and strong. However, if volume and price are changing in different directions, a reversal might be happened.

You can analyze it in the previous picture.

Price - Moving Average

In terms of trading price, traders often observed the trends based on the charts shape and cross in ways that form shapes - often times with weird names like ‘head and shoulder’, ’ reverse head and sholder’, ‘double top’, ‘golden cross’, etc. A golden cross indicates a long term bull market going forward, whereas the death cross is the exact opposite, indicating a potential long term bear market. Both of these refer to the confirmation of long term trend by the occurance of the overlapping of moving average lines as shown below.

In the quantmod package in R, the addMA function can be used to add moving averages to a financial time series chart. There are various methods of calculating moving averages that can be used in this function, which I will describe below:

1. Simple Moving Average (SMA): This is the most basic type of moving average and is calculated by taking the average of a set number of periods of data. For example, a 20-day SMA would be the average of the closing prices for the last 20 days. This method gives equal weight to all periods in the moving average. \[SMA_t = \frac{P_t + \dots + P_{t-n+1}}{n}\] Buy signal arises when a short-run SMA crosses from below to above a long-run SMA. Sell signal arrises when a short-run SMA crosses from above to below a long-run SMA.

2. Weighted Moving Average (WMA): This method assigns a weight to each data point in the moving average, with more recent data points given greater weight. The weights can be calculated in various ways, such as linear or exponential.

3. Exponential Moving Average (EMA): This method is similar to the WMA, but it gives greater weight to more recent data points using an exponential decay formula. The formula used to calculate the EMA involves a smoothing factor, which determines the weight given to each data point. \[EMA_t = \beta P_t + (1-\beta) EMA_{t-1}\] where the smoothing coefficient is usually defined \(\beta = \frac{2}{n+1} \in (0;1)\).

4. Double Exponential Moving Average (DEMA): This method is similar to the EMA, but it uses a second EMA to smooth the first EMA. This results in a smoother moving average that is less sensitive to short-term fluctuations.

5. Triple Exponential Moving Average (TEMA): This method is similar to the DEMA, but it uses a third EMA to further smooth the moving average. This method is even less sensitive to short-term fluctuations than the DEMA.

These are the main types of moving averages that can be used in the addMA function in quantmod. Each method has its own advantages and disadvantages, and the choice of method will depend on the specific needs of the analysis being performed.

chartSeries(stock,
            subset='2020-01-01::2020-09-23', type = "candlestick",
            theme=chartTheme('white'))

addSMA(n = 12, on = 1, with.col = Cl, overlay = TRUE, col = "brown")

addSMA(n = 24, on = 1, with.col = Cl, overlay = TRUE, col = "blue")

NA
NA

in January 02 the trend was very high on the graph, in March 02 it started to decrease dramatically, and then from May01 until September01 it shows a slow increase ## Moving Average Convergence Divergence (MACD)

chartSeries(stock,
            subset='2020-01-01::2020-09-23', type = "bars",
            theme=chartTheme('white'))

addMACD(fast=12,slow=26,signal=9,type="EMA")

the chart shows a high trend from january 02 until march 02, then a decrease from april 01. after that a continuous slow increase from barley until september 23