RSI

RSI Visualization

library(quantmod)

## 필요한 패키지를 로딩중입니다: xts

## 필요한 패키지를 로딩중입니다: zoo

## 
## 다음의 패키지를 부착합니다: 'zoo'

## The following objects are masked from 'package:base':
## 
##     as.Date, as.Date.numeric

## 필요한 패키지를 로딩중입니다: TTR

## Registered S3 method overwritten by 'quantmod':
##   method            from
##   as.zoo.data.frame zoo

library(TTR)
library(highcharter)

# Specify the ticker symbol and the source
symbol <- "AAPL"

# Import the historical stock prices
aapl <- getSymbols(symbol, src ='yahoo',from='2022-01-01')


# Extract the adjusted close prices
prices <- Cl(get(symbol))

# Calculate the RSI
rsi <- RSI(prices)

#Use highcharter for visualization
hc1 <- hchart(AAPL) %>%
  hc_title(text = "Daily Stock Price of AAPL (2022M1 - 2023M3)") %>%
  hc_xAxis(type = "datetime") %>%
  hc_yAxis(title = list(text = "Price"))
hc2 <- hchart(rsi) %>%
  hc_title(text = "RSI of AAPL (2022M1 - 2023M3)") %>%
  hc_xAxis(type = "datetime") %>%
  hc_yAxis(title = list(text = "RSI"))

grid <- hw_grid(list(hc1, hc2),ncol=1) #Plot two graphs in one grid
grid

library(ggplot2)
library(zoo)
library(dplyr)

## 
## 다음의 패키지를 부착합니다: 'dplyr'

## The following objects are masked from 'package:xts':
## 
##     first, last

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(DT)

#dataframe for RSI and Close price
rsi_df <- data.frame(date = index(rsi), rsi = coredata(rsi),close=as.numeric(AAPL$AAPL.Close))
datatable(rsi_df)

# Plot RSI with shaded areas
gg1 <- ggplot(rsi_df, aes(x = date, y = rsi)) +
  geom_line(size=1, color="#5F9EA0") +
  geom_rect(aes(xmin = date, xmax = lead(date), ymin = ifelse(rsi < 30, -Inf, NA), ymax = Inf), fill = "blue", alpha = 0.2) +
  geom_rect(aes(xmin = date, xmax = lead(date), ymin = -Inf, ymax = ifelse(rsi > 60, Inf, NA)), fill = "red", alpha = 0.2,inherit.aes = FALSE) +
  scale_y_continuous(breaks=seq(0,100,10)) +
  labs(title=paste("RSI for", symbol), x="Date", y="RSI") +
  scale_x_date(date_breaks = '2 months',
               date_labels = '%Y-%m') +
  theme_light()

## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## i Please use `linewidth` instead.

# Plot the stock price with shaded areas
gg2 <- ggplot(rsi_df, aes(x = date, y = close)) +
  geom_line(size=1, color="cornflowerblue") +
  geom_rect(aes(xmin = date, xmax = lead(date), ymin = ifelse(rsi < 30, -Inf, NA), ymax = Inf), fill = "blue", alpha = 0.2) +
  geom_rect(aes(xmin = date, xmax = lead(date), ymin = -Inf, ymax = ifelse(rsi > 60, Inf, NA)), fill = "red", alpha = 0.2) +
  
  labs(title=paste("Close Price for", symbol), x="Date", y="Close Price") +
  scale_x_date(date_breaks = '2 months',
               date_labels = '%Y-%m') +
  theme_light()

library(patchwork) 
gg2+gg1+plot_layout(ncol=1) #Helps plot two graphs in one grid for ggplot2 objects

## Warning: Removed 298 rows containing missing values (`geom_rect()`).

## Warning: Removed 251 rows containing missing values (`geom_rect()`).

## Warning: Removed 14 rows containing missing values (`geom_line()`).

## Warning: Removed 298 rows containing missing values (`geom_rect()`).

## Warning: Removed 251 rows containing missing values (`geom_rect()`).

The blue area shows where the RSI is less than 30 and the red area shows where the RSI is over 60. In this exercise we used 60 as the standard not 70 considering that the stock market was in correction since last year.

RSI Trading

For this part, we are going to implement RSI trading strategy. We are going to buy the stock when the RSI below 30 recovers and goes above 30 and sell it when RSI above 60 goes down below 60, and then repeat this process again until we find all of the trading points.

# Finding indices for buying and selling 
# Define threshold values
buy_threshold <- 30
sell_threshold <- 60

# Initialize variables
position <- "short"
last_signal <- "none"

buy_signals <- c()
sell_signals <- c()



# Loop through RSI values
for (i in 1:length(rsi)) {
  # Check for buy signal
  if (!is.na(rsi[i]) && rsi[i] < buy_threshold && position == "short") {
    # Wait for RSI to cross above buy threshold
    for (j in (i+1):length(rsi)) {
      if (!is.na(rsi[j]) && rsi[j] > buy_threshold) {
        position <- "long"
        last_signal <- "buy"
        buy_signals <- c(buy_signals, j)
        break  # Exit loop once buy signal is found
      }
    }
  }
  
  # Check for sell signal
  if (!is.na(rsi[i]) && rsi[i] > sell_threshold && position == "long") {
    # Wait for RSI to cross below sell threshold
    for (j in (i+1):length(rsi)) {
      if (!is.na(rsi[j]) && rsi[j] < sell_threshold) {
        position <- "short"
        last_signal <- "sell"
        sell_signals <- c(sell_signals, j)
        
        # Find all buy signals between the current and next sell signals
        for (k in (buy_signals[length(buy_signals)]+1):(j-1)) {
          if (!is.na(rsi[k]) && rsi[k] < buy_threshold) {
            # Wait for RSI to cross above buy threshold
            for (m in (k+1):j) {
              if (!is.na(rsi[m]) && rsi[m] > buy_threshold) {
                # Buy signal found between sell signals
                buy_signals <- c(buy_signals, m)
                break
              }
            }
          }
        } 
        break  # Exit loop once sell signal is found
      }
    }
  }
}

#Buy signal Index
buy_signals

## [1]  19 189 250

#Sell Signal Index
sell_signals

## [1]  62 285

# Create data frame with Date and Close Price columns
nf <- data.frame(Date = index(rsi), Close = rsi_df$close)

# Add Trade column and initialize to "none"
nf$Trade <- rep(NA, nrow(nf))

# Set buy signals in Trade column to "buy"
nf$Trade[buy_signals] <- "buy"

# Set sell signals in Trade column to "sell"
nf$Trade[sell_signals] <- "sell"

# Sort data frame by Date
nf <- nf[order(nf$Date),]
nf_omit <- na.omit(nf)
datatable(nf_omit)

The table above shows when we have to buy the stocks and sell the stocks according to the strategy

library(ggplot2)


# Filter out NA trades
trades <- na.omit(nf)

# Plot RSI and trades
gg_rsi1 <- ggplot(nf, aes(x = Date, y = Close)) +
  geom_line(size=0.8, color="cornflowerblue") +
  geom_point(data = trades[trades$Trade == "buy", ], aes(color = "buy"), size = 4) +
  geom_point(data = trades[trades$Trade == "sell", ], aes(color = "sell"), size = 4) +
  scale_color_manual(values = c("buy" = "grey30", "sell" = "orangered")) +
  labs(x = "Date", y = "Price",color='Signal') +
  ggtitle("RSI Trading Signal (AAPL)") + 
  scale_x_date(date_breaks = '2 months',
               date_labels = '%Y-%m') +
  theme_light()+
  theme(legend.position = "top") 

gg_rsi1

#Calculate Returns

# Filter out NA trades
trades <- na.omit(nf)

# For first trade
buy_price <- trades$Close[1]
sell_price <- trades$Close[2]
returns1 <- (sell_price - buy_price) / buy_price

# For second trade
buy_prices <- c(trades$Close[3], trades$Close[4])
sell_price <- trades$Close[5]
returns2 <- (sell_price - mean(buy_prices)) / mean(buy_prices)


# Total Cumulative Return
trading_return <- (1+returns1)*(1+returns2)-1

# Buy&Hold
buy_hold_return <- (nf[301,'Close']-nf[1,'Close'])/nf[1,'Close']

#Make a dataframe to plot the return graph
returns_df <- data.frame(
  Trade = c("1st Trade", "2nd Trade", "RSI Trading Profit",'Buy&Hold'),
  Return = c(returns1, returns2,trading_return,buy_hold_return)
)

# Define the order of the trades
trade_order <- c("1st Trade", "2nd Trade", "RSI Trading Profit", "Buy&Hold")
returns_df$Trade <- factor(returns_df$Trade, levels = trade_order)

gg1 <- ggplot(returns_df, aes(x=Trade, y=Return,color=Trade,fill=Trade)) +
  geom_bar(stat="identity",color='black',width = 0.5 ) +
  labs(title="Returns from RSI Trading Strategy for AAPL", y="Return", x="Trade") +
  scale_y_continuous(labels = scales::percent)+
  theme_light()+
  geom_text(aes(label=scales::percent(Return)), vjust=-0.2, size=4)+
  geom_text(data = filter(returns_df, Return < 0), aes(label=scales::percent(Return), y=Return - 0.01), size=4)+
  theme(legend.position = "top") 
gg1

RSI vs Bollinger Band

We see that the rather than buy&holding the stock, trading made a better outcome. Now, let’s compare the performance with the Bollinger Band strategy.

AAPL$BBands <- BBands(Cl(AAPL), n = 20, sd = 2)

#Add trade signal
AAPL$signal_b <- ifelse(Cl(AAPL) < AAPL$dn, 1, #If the close prices is below the band, it's a buy signal
                 ifelse(Cl(AAPL) > AAPL$up, -1, 0)) #If the close prices is above the band, it's a sell signal

#Create trade signal using bollingerband strategy
trade_b <- ifelse(AAPL$signal == 1,"buy",
                ifelse(AAPL$signal == -1,"sell",""))

#Create new dataframe for trading strategy
df_b <- data.frame(Date=index(AAPL), Close=Cl(AAPL), BBands_dn = AAPL$dn, BBands_up = AAPL$up,Trade=trade_b)
#Change the column names
colnames(df_b) <- c('Date','Close','BBands_dn','BBands_up','Trade')

datatable(df_b)

#Visualize 
gg_bband1 <- ggplot(df_b, aes(x=Date, y=Close)) +
  geom_line(size=0.8,color="steelblue") + 
  geom_ribbon(aes(ymin=BBands_dn, ymax=BBands_up), alpha=0.2, fill="black") +
  geom_point(data=df_b[df_b$Trade=="buy",], aes(x=Date, y=Close, color="buy"), size=3) +
  geom_point(data=df_b[df_b$Trade=="sell",], aes(x=Date, y=Close, color="sell"), size=3) +
  scale_color_manual(name="Trade", values=c("buy"="grey30", "sell"="orangered")) +
  labs(title="Bollinger Band Signal (AAPL)", y="Price", x="Date")+
  theme_light()+
  theme(legend.position = "top") +
  scale_x_date(date_breaks = '2 months',
               date_labels = '%Y-%m') 

library(patchwork)
gg_bband1 +gg_rsi1 + plot_layout(ncol=1)

## Warning: Removed 19 rows containing missing values (`geom_point()`).
## Removed 19 rows containing missing values (`geom_point()`).

#Identify the first trade relevant indices
first_buy_index <- which(df_b$Trade=="buy")[1]
first_sell_index <- which(df_b$Trade=="sell" & index(df_b) > index(df_b)[first_buy_index])[1]

#Identify the second trade relevant indices
second_buy_index <- which(df_b$Trade=="buy" & index(df_b) > index(df_b)[first_sell_index])[1]
second_sell_index <- which(df_b$Trade=="sell" & index(df_b) > index(df_b)[second_buy_index])[1]

#Identify the third trade relevant indices
third_buy_index <- which(df_b$Trade =='buy' & index(df_b) > index(df_b)[second_sell_index])[1]
third_sell_index <- which(df_b$Trade=='sell'& index(df_b)> index(df_b)[third_buy_index])[1]

#Identify the fourth trade relevant indices
fourth_buy_index <- which(df_b$Trade =='buy' & index(df_b) > index(df_b)[third_sell_index])[1]
fourth_sell_index <- which(df_b$Trade=='sell'& index(df_b)> index(df_b)[fourth_buy_index])[1]

# Calculate average price and return for first buy-sell pair
if(!is.na(first_buy_index) & !is.na(first_sell_index)){
  buy_prices_1 <- df_b[which(df_b$Trade == "buy" & index(df_b) <= index(df_b)[first_sell_index]), "Close"]
  buy_avg_price_1 <- mean(buy_prices_1, na.rm = TRUE)
  
  sell_price_1 <- df_b[first_sell_index, "Close"]
  return_1 <- (sell_price_1 - buy_avg_price_1) / buy_avg_price_1
} else {
  return_1 <- NA
}

# Calculate average price and return for second buy-sell pair
if(!is.na(second_buy_index) & !is.na(second_sell_index)){
  buy_prices_2 <- df_b[which(df_b$Trade == "buy" & index(df_b) > index(df_b)[first_sell_index] & 
                          index(df_b) < index(df_b)[second_sell_index]), "Close"]
  buy_avg_price_2 <- mean(buy_prices_2, na.rm = TRUE)
  
  sell_price_2 <- df_b[second_sell_index, "Close"]
  return_2 <- (sell_price_2 - buy_avg_price_2) / buy_avg_price_2
} else {
  return_2<- NA
}

# Calculate average price and return for third buy-sell pair
if(!is.na(third_buy_index) & !is.na(third_sell_index)){
  buy_prices_3 <- df_b[which(df_b$Trade == "buy" & index(df_b) > index(df_b)[second_sell_index] & 
                          index(df_b) < index(df_b)[third_sell_index]), "Close"]
  buy_avg_price_3 <- mean(buy_prices_3, na.rm = TRUE)
  
  sell_price_3 <- df_b[third_sell_index, "Close"]
  return_3 <- (sell_price_3 - buy_avg_price_3) / buy_avg_price_3
} else {
  return_3<- NA
}


# Calculate average price and return for fourth buy-sell pair
if(!is.na(fourth_buy_index) & !is.na(fourth_sell_index)){
  buy_prices_4 <- df_b[which(df_b$Trade == "buy" & index(df_b) > index(df_b)[third_sell_index] & 
                          index(df_b) < index(df_b)[fourth_sell_index]), "Close"]
  buy_avg_price_4 <- mean(buy_prices_4, na.rm = TRUE)
  
  sell_price_4 <- df_b[fourth_sell_index, "Close"]
  return_4 <- (sell_price_4 - buy_avg_price_4) / buy_avg_price_4
} else {
  return_4<- NA
}



bollinger_return <- (1 + return_1) * (1 + return_2) * (1 + return_3) * (1 + return_4)  - 1

#Create Dataframe for visualize trade profit
returns_df <- data.frame(
  Trade = c("1st Trade", "2nd Trade", "3rd Trade", "4th Trade", "Bollinger Trading Profit",'Buy&Hold'),
  Return = c(return_1, return_2, return_3, return_4, bollinger_return,buy_hold_return)
)
trade_order <- c("1st Trade", "2nd Trade", "3rd Trade", "4th Trade", "Bollinger Trading Profit", "Buy&Hold")
returns_df$Trade <- factor(returns_df$Trade, levels = trade_order)

# plot the returns using a bar chart
gg2 <- ggplot(returns_df, aes(x=Trade, y=Return,color=Trade, fill=Trade)) +
  geom_bar(stat="identity",color='black' ) +
  labs(title="Returns from Bollinger Band Trading Strategy for AAPL", y="Return", x="Trade") +
  scale_y_continuous(labels = scales::percent)+
  geom_text(aes(label=scales::percent(Return)), vjust=-0.2, size=4)+
  geom_text(data = filter(returns_df, Return < 0), aes(label=scales::percent(Return), y=Return - 0.01), size=4)+
  theme_light()+
  theme(legend.position = "top") 


gg2+gg1+patchwork::plot_layout(ncol=1)

#Make a dataframe to plot the return graph
returns_df <- data.frame(
  Trade = c("RSI Trading Profit","Bollinger Trading Profit", 'Buy&Hold'),
  Return = c(trading_return,bollinger_return,buy_hold_return)
)

# Define the order of the trades
trade_order <- c("RSI Trading Profit","Bollinger Trading Profit", 'Buy&Hold')
returns_df$Trade <- factor(returns_df$Trade, levels = trade_order)

returns_df

##                      Trade     Return
## 1       RSI Trading Profit  0.1189514
## 2 Bollinger Trading Profit  0.3131879
## 3                 Buy&Hold -0.1594417

ggplot(returns_df, aes(x=Trade, y=Return, fill=Trade)) +
  geom_bar(stat="identity", color='black', width=0.5 ) +
  labs(title="Trading Strategy Return (AAPL)", y="Return", x="Trade") +
  scale_y_continuous(labels = scales::percent) +
  geom_text(aes(label=scales::percent(Return)), vjust=-0.2, size=4) +
  scale_fill_manual(values = c("RSI Trading Profit" = "cornflowerblue",
                               "Bollinger Trading Profit" = "#A29BFE",
                               "Buy&Hold" = "lightgreen")) +
  
  theme_light() +
  theme(legend.position = "top")