# Load packages
# Core
library(tidyverse)
library(tidyquant)
Measure portfolio risk using skewness. Skewness is the extent to which returns are asymmetric around the mean. It is important because a positively skewed distribution means large positive returns are more likely while a negatively skewed distribution implies large negative returns are more likely.
five stocks: “SPY”, “EFA”, “IJS”, “EEM”, “AGG”
from 2012-12-31 to 2017-12-31
symbols <- c("SPY", "EFA", "IJS", "EEM", "AGG")
prices <- tq_get(x = symbols,
get = "stock.prices",
from = "2012-12-31",
to = "2017-12-31")
asset_returns_tbl <- prices %>%
group_by(symbol) %>%
tq_transmute(select = adjusted,
mutate_fun = periodReturn,
period = "monthly",
type = "log") %>%
slice(-1) %>%
ungroup() %>%
set_names(c("asset", "date", "returns"))
# symbols
symbols <- asset_returns_tbl %>% distinct(asset) %>% pull()
symbols
## [1] "AGG" "EEM" "EFA" "IJS" "SPY"
# weights
weights <- c(0.25, 0.25, 0.2, 0.2, 0.1)
weights
## [1] 0.25 0.25 0.20 0.20 0.10
w_tbl <- tibble(symbols, weights)
w_tbl
## # A tibble: 5 × 2
## symbols weights
## <chr> <dbl>
## 1 AGG 0.25
## 2 EEM 0.25
## 3 EFA 0.2
## 4 IJS 0.2
## 5 SPY 0.1
# ?tq_portfolio
portfolio_returns_tbl <- asset_returns_tbl %>%
tq_portfolio(assets_col = asset,
returns_col = returns,
weights = w_tbl,
rebalance_on = "months",
col_rename = "returns")
portfolio_returns_tbl
## # A tibble: 60 × 2
## date returns
## <date> <dbl>
## 1 2013-01-31 0.0204
## 2 2013-02-28 -0.00239
## 3 2013-03-28 0.0121
## 4 2013-04-30 0.0174
## 5 2013-05-31 -0.0128
## 6 2013-06-28 -0.0247
## 7 2013-07-31 0.0321
## 8 2013-08-30 -0.0224
## 9 2013-09-30 0.0511
## 10 2013-10-31 0.0301
## # ℹ 50 more rows
portfolio_returns_tbl %>%
tq_performance(Ra = returns,
Rb = NULL,
performance_fun = table.Stats) %>%
select(Skewness)
## # A tibble: 1 × 1
## Skewness
## <dbl>
## 1 -0.168
# Figure 5.2 Shaded histogram returns ----
portfolio_returns_tbl %>%
# Create a new variable for shade
mutate(returns_extreme_neg = if_else(returns < mean(returns) - 2*sd(returns),
"yes",
"no")) %>%
# Plot
ggplot(aes(returns, fill = returns_extreme_neg)) +
geom_histogram(alpha = .7,
binwidth = .003) +
scale_x_continuous(breaks = scales::pretty_breaks(n = 8)) +
scale_fill_tq() +
labs(x = "monthly returns")
# Figure 5.6 Asset and portfolio skewness comparison ----
asset_returns_skew_tbl <- asset_returns_tbl %>%
# skewness for each asset
group_by(asset) %>%
summarise(skew = skewness(returns)) %>%
ungroup() %>%
# skewness of portfolio
add_row(tibble(asset = "Portfolio",
skew = skewness(portfolio_returns_tbl$returns)))
asset_returns_skew_tbl %>%
ggplot(aes(asset, skew, color = asset)) +
geom_point() +
# Add label for portfolio
ggrepel::geom_text_repel(aes(label = asset),
data = asset_returns_skew_tbl %>%
filter(asset == "Portfolio"),
size = 5,
show.legend = FALSE) +
labs(y = "skewness")
# 3 Rolling skewness ----
# Why rolling skewness?
# To check anything unusual in the portfolio's historical risk
# Assign a value to winder
window <- 24
port_rolling_sd_tbl <- portfolio_returns_tbl %>%
tq_mutate(select = returns,
mutate_fun = rollapply,
width = window,
FUN = skewness,
col_rename = "rolling_skew") %>%
select(date, rolling_skew) %>%
na.omit()
# Figure 4.8 Rolling skewness ggplot ----
port_rolling_sd_tbl %>%
ggplot(aes(date, rolling_skew)) +
geom_line(color = "cornflowerblue") +
geom_hline(yintercept = 0, linetype = "dotted", size = 2) +
scale_y_continuous(limits = c(-1,1),
breaks = scales::pretty_breaks(n = 10)) +
scale_x_date(breaks = scales::breaks_pretty(n = 7))+
labs(title = paste0("Rolling ", window, "-Month Skew"),
x = NULL,
y = "skewness") +
theme(plot.title = element_text(hjust = 0.5)) +
annotate(geom = "text",
x = as.Date("2016-09-01"), y = 0.7,
color = "red", size = 5,
label = str_glue("The 24-month skewness is positive for about half of the lifetime,
even though the overall skewness is negative"))