In frdanconia/shiny_etf_portfolio: ShinyETF

In a previous post, we created an R Notebook to explore the relationship between the copper/gold price ratio and 10-year Treasury yields. Today, we'll create a Shiny app that lets the user choose among a choice of ratios and see how they correlate with a variety of other indicators. For example, perhaps a user doesn't care about Dr. Copper, but instead wants to explore the oil/gold price ratio and how it correlates with the US inflation rate, or the EU inflation rate.

The finished app is available here.

# We are choosing a few commodities and  
# Remember this is a commodities flexdashboard. 

commodityChoices <- c(
                  "Copper" = "CHRIS/CME_HG1",
                  "WTI oil" = "FRED/DCOILWTICO",# 
                  "Iron Ore" = "ODA/PIORECR_USD", # monthly
                  "Platinum" = "LPPM/PLAT", 
                  "Palladium" = "LPPM/PALL",
                  "Silver" = "LBMA/SILVER") 

selectInput("commodity",
            "Commodity",
            choices = commodityChoices, 
            selected = "Copper")

Now let's give the users a choice of the economic indicator correlation to test/visualize.

# Let's choose a few economic indicators. There are others, like 
# unemployment, consumer confidence, home building, retail spending.
econIndicatorChoices <- c(
                  "10-Yr Yield" = "FRED/DGS10", # daily
                  "US CPI" = "RATEINF/INFLATION_USA",# monthly
                  "Japan CPI" = "RATEINF/INFLATION_JPN",
                  "EU CPI" = "RATEINF/INFLATION_EUR") 

selectInput("econIndicator",
            "Economic Indicator",
            choices = econIndicatorChoices, 
            selected = "10-yr Yield")


dateRangeInput("dateRange",
               "Date range",
               start = "1990-01-01",
               end   = "2016-12-31")

Now that we have the inputs, we need to import the data for the chosen commodity, gold and the chosen economic indicator. Then, we create ratio of commodity/gold prices and save it in an object with the economic indicator time series.

ratio_indicator <- reactive({

# We have to pass the start and end dates to Quandl three times, 
# so let's save them in nicer object. 
start_date <- format(input$dateRange[1])
end_date <- format(input$dateRange[2])

# First let's import the price data for the commodity chosen by the user.    
commodity <- Quandl(input$commodity,
               start_date = start_date,
               end_date = end_date,
               order = "asc",
               type = "xts")

# And we are not giving the user a choice for the denominator. It is 
# gold prices. We could offer up a choice if we think there's another
# safe-haven asset that could lead to better or more interesting
# insights. But for today it's  gold, Jerry, gold. 

gold <- Quandl("CHRIS/CME_GC1",
               start_date = start_date,
               end_date = end_date, 
               order = "asc",
               type = "xts")

# Now let's import the price data for the econimic indicator chosen by the user.

econIndicator <- Quandl(input$econIndicator,
               start_date = start_date,
               end_date = end_date,
               order = "asc",
               type = "xts")

Okay, we have our base data now: thet time series for the commodity, the time series for gold, and the time series for the economic indicator. Now we need to calcuate the price ratio of commodity/gold.

First we will merge the data with merge.xts() and replace NAs as we did in the original Notebook. We will use the na.locf() function for this.

# Merge the time series that we just imported from Quandl. 
data_merged <- na.locf(merge.xts(commodity[,1], gold$Settle, econIndicator), formLast = TRUE)
# Add better column names.
colnames(data_merged ) <- c(names(commodityChoices[commodityChoices == input$commodity]), 
                            "Gold",
                            names(econIndicatorChoices[econIndicatorChoices == input$econIndicator]))

Now we need to create a ratio - this is why we had to replace those NAs - as a new column in the xts object. It's not necessary but just for clarity we can go ahead and save just the ratio and the economic indicator time series.

# Create the ratio in a new column
data_merged$ratio <- (data_merged[,1])/data_merged$Gold

# Save just the ratio and the gold time series.

ratio_indicator  <- merge(data_merged$ratio, data_merged[,3])

colnames(ratio_indicator) <- c("ratio","indicator")

# Return our finished xts object.
return(ratio_indicator)
})

We now have a reactive called ratio_indicator()

Chart 1

dygraphOutput("ratio_indicator")

output$ratio_indicator <- renderDygraph({
  dygraph(ratio_indicator()) %>% 
  # Add the rollPeriod for smoothing.
  dyRoller(rollPeriod = 3) %>% 
  # Create two independent axes.
  dyAxis("y", label = "USD") %>%
  dyAxis("y2", label = "Percent (%)", independentTicks = TRUE) %>%
  # Assign each time series to an axis.
  dySeries("ratio", axis = 'y', 
           label = paste(names(commodityChoices[commodityChoices == input$commodity]), 
                         "/Gold (LHS)", sep = ""), color = "blue") %>% 
  dySeries("indicator", axis = 'y2', 
           label = paste(names(econIndicatorChoices[econIndicatorChoices == input$econIndicator]), 
                         "(RHS)", sep = ""), color = "red")
})

Row

Chart 2

dygraphOutput("rollingCorrelation")

output$rollingCorrelation <- renderDygraph({

rolling_cor <- rollapply(ratio_indicator(), 24,
                         function(x) cor(x[, 1], x[, 2], use = "pairwise.complete.obs"),
                         by.column = FALSE)

names(rolling_cor) <- paste(names(commodityChoices[commodityChoices == input$commodity]),
                            "/Gold ",
                            names(econIndicatorChoices[econIndicatorChoices == input$econIndicator]), 
                            " Correlation", sep = "")

  avg <- round(mean(rolling_cor,  na.rm = T), 2)
  mini <- round(min(rolling_cor,  na.rm = T), 2)

  maxi <- round(max(rolling_cor,  na.rm = T), 2)


dygraph(rolling_cor, main = paste(names(commodityChoices[commodityChoices == input$commodity]),
                            "/Gold ",
                            names(econIndicatorChoices[econIndicatorChoices == input$econIndicator]), 
                            " Correlation", sep = "")) %>% 
  dyRangeSelector(dateWindow = c("2015-01-01", "2016-12-31"))  %>% 
  dyLimit(avg, color = 'purple') %>% 
  dyLimit(mini, color = 'red') %>% 
  dyLimit(maxi, color = 'blue') %>% 
  dyEvent("2016-11-08", "Trump!", labelLoc = "bottom")
})