Endogenously Detecting Structural Breaks in a Time Series: Implementation in R

The most conventional approach to determine structural breaks in longitudinal data seems to be the Chow Test.

From Wikipedia,

The Chow test, proposed by econometrician Gregory Chow in 1960, is a test of whether the coefficients in two linear regressions on different data sets are equal. In econometrics, it is most commonly used in time series analysis to test for the presence of a structural break at a period which can be assumed to be known a priori (for instance, a major historical event such as a war). In program evaluation, the Chow test is often used to determine whether the independent variables have different impacts on different subgroups of the population.

As shown in the figure below, regressions on the 2 sub-intervals seem to have greater explanatory power than a single regression over the data.


For the data above, determining the sub-intervals is an easy task. However, things may not look that simple in reality. Conducting a Chow test for structural breaks leaves the data scientist at the mercy of his subjective gaze in choosing a null hypothesis for a break point in the data.

Instead of choosing the breakpoints in an exogenous manner, what if the data itself could learn where these breakpoints lie? Such an endogenous technique is what Bai and Perron came up with in a seminal paper published in 1998 that could detect multiple structural breaks in longitudinal data. A later paper in 2003 dealt with the testing for breaks empirically, using a dynamic programming algorithm based on the Bellman principle.

I will discuss a quick implementation of this technique in R.

Brief Outline:

Assuming you have a ts object (I don’t know whether this works with zoo, but it should) in R, called ts. Then implement the following:

# assuming you have a 'ts' object in R
# 1. install package 'strucchange'
# 2. Then write down this code:
# store the breakdates
bp_ts <- breakpoints(ts ~ 1)
# this will give you the break dates and their confidence intervals
# store the confidence intervals
ci_ts <- confint(bp_ts)
## to plot the breakpoints with confidence intervals

An illustration 

I started with data on India’s rice crop productivity between 1950 (around Independence from British Colonial rule) and 2008. Here’s how it looks:


You can download the excel and CSV files here and here respectively.

Here’s the way to go using R:

## load the data from a CSV or Excel file. This example is done with an Excel sheet.
prod_df <- read.xlsx(file = 'agricultural_productivity.xls', sheetIndex = 'Sheet1', rowIndex = 8:65, colIndex = 2, header = FALSE)
colnames(prod_df) <- c('Rice')
## store rice data as time series objects
rice <- ts(prod_df$Rice, start=c(1951, 1), end=c(2008, 1), frequency=1)
# store the breakpoints
bp.rice <- breakpoints(rice ~ 1)
## the BIC chooses 5 breakpoints; plot the graph with breakdates and their confidence intervals
## confidence intervals
ci.rice <- confint(bp.rice)

Voila, this is what you get:


The dotted vertical lines indicated the break dates; the horizontal red lines indicate their confidence intervals.

This is a quick and dirty implementation. For a more detailed take, check out the documentation on the R package called strucchange.


Detecting Structural Breaks in China’s FX Regime

Edit: This post is in its infancy. Work is still ongoing as far as deriving insight from the data is concerned. More content and economic insight is expected to be added to this post as and when progress is made in that direction.

This is an attempt to detect structural breaks in China’s FX regime using Frenkel Wei regression methodology (this was later improved by Perron and Bai). I came up with the motivation to check for these structural breaks while attending a guest lecture on FX regimes by Dr. Ajay Shah delivered at IGIDR. This is work that I and two other classmates are working on as a term paper project under the supervision of Dr. Rajeswari Sengupta.

The code below can be replicated and run as is, to get same results.

## if fxregime or strucchange package is absent from installed packages, download it and load it
## load packages
# load the necessary data related to exchange rates - 'FXRatesCHF'
# this dataset treats CHF as unit currency
data("FXRatesCHF", package = "fxregime")
## compute returns for CNY (and explanatory currencies)
## since China abolished fixed USD regime
cny <- fxreturns("CNY", frequency = "daily",
start = as.Date("2005-07-25"), end = as.Date("2010-02-12"),
other = c("USD", "JPY", "EUR", "GBP"))
## compute all segmented regression with minimal segment size of
## h = 100 and maximal number of breaks = 10
regx <- fxregimes(CNY ~ USD + JPY + EUR + GBP,
data = cny, h = 100, breaks = 10, ic = "BIC")
## Print summary of regression results
## minimum BIC is attained for 2-segment (1-break) model
round(coef(regx), digits = 3)
sqrt(coef(regx)[, "(Variance)"])
## inspect associated confidence intervals
cit <- confint(regx, level = 0.9)
## plot LM statistics along with confidence interval
flm <- fxlm(CNY ~ USD + JPY + EUR + GBP, data = cny)
scus <- gefp(flm, fit = NULL)
plot(scus, functional = supLM(0.1))
## add lines related to breaks to your plot

As can be seen in the figure below, the structural breaks correspond to the vertical bars. We are still working on understanding the motivations of China’s central bank in varying the degree of the managed float exchange rate.


EDIT (May 16, 2016):

The code above uses data provided by the package itself. If you wished to replicate this analysis on data after 2010, you will have to use your own data. We used Quandl, which lets you get 10 premium datasets for free. An API key (for only 10 calls on premium datasets) is provided if you register there. Foreign exchange rate data (2000 onward till date) apparently, is premium data. You can find these here.

Here are the (partial) results and code to work the same methodology on the data from 2010 to 2016:


## if fxregime is absent from installed packages, download it and load it
## load package
# load the necessary data related to exchange rates - 'FXRatesCHF'
# this dataset treats CHF as unit currency
# install / load Quandl
# Extract and load currency data series with respect to CHF from Quandl
# Extract data series from Quandl. Each Quandl user will have unique api_key
# upon signing up. The freemium version allows access up to 10 fx rate data sets
# USDCHF <- Quandl("CUR/CHF", api_key="p2GsFxccPGFSw7n1-NF9")
# write.csv(USDCHF, file = "USDCHF.csv")
# USDCNY <- Quandl("CUR/CNY", api_key="p2GsFxccPGFSw7n1-NF9")
# write.csv(USDCNY, file = "USDCNY.csv")
# USDJPY <- Quandl("CUR/JPY", api_key="p2GsFxccPGFSw7n1-NF9")
# write.csv(USDJPY, file = "USDJPY.csv")
# USDEUR <- Quandl("CUR/EUR", api_key="p2GsFxccPGFSw7n1-NF9")
# write.csv(USDEUR, file = "USDEUR.csv")
# USDGBP <- Quandl("CUR/GBP", api_key="p2GsFxccPGFSw7n1-NF9")
# write.csv(USDGBP, file = "USDGBP.csv")
# load the data sets into R
USDCHF <- read.csv("G:/China's Economic Woes/USDCHF.csv")
USDCNY <- read.csv("G:/China's Economic Woes/USDCNY.csv")
USDEUR <- read.csv("G:/China's Economic Woes/USDEUR.csv")
USDGBP <- read.csv("G:/China's Economic Woes/USDGBP.csv")
USDJPY <- read.csv("G:/China's Economic Woes/USDJPY.csv")
start = 1 # corresponds to 2016-05-12
end = 2272 # corresponds to 2010-02-12
dates <- as.Date(USDCHF[start:end,1])
USD <- 1/USDCHF[start:end,2]
CNY <- USDCNY[start:end,2]/USD
JPY <- USDJPY[start:end,2]/USD
EUR <- USDEUR[start:end,2]/USD
GBP <- USDGBP[start:end,2]/USD
# reverse the order of the vectors to reflect dates from 2005 - 2010 instead of
# the other way around
USD <- USD[length(USD):1]
CNY <- CNY[length(CNY):1]
JPY <- JPY[length(JPY):1]
EUR <- EUR[length(EUR):1]
GBP <- GBP[length(GBP):1]
dates <- dates[length(dates):1]
df <- data.frame(CNY, USD, JPY, EUR, GBP)
df$weekday <- weekdays(dates)
row.names(df) <- dates
df <- subset(df, weekday != 'Sunday')
df <- subset(df, weekday != 'Saturday')
df <- df[,1:5]
zoo_df <- as.zoo(df)
# Code to replicate analysis
cny_rep <- fxreturns("CNY", data = zoo_df, frequency = "daily",
other = c("USD", "JPY", "EUR", "GBP"))
time(cny_rep) <- as.Date(row.names(df)[2:1627])
regx_rep <- fxregimes(CNY ~ USD + JPY + EUR + GBP,
data = cny_rep, h = 100, breaks = 10, ic = "BIC")
## minimum BIC is attained for 2-segment (5-break) model
round(coef(regx_rep), digits = 3)
sqrt(coef(regx_rep)[, "(Variance)"])
## inspect associated confidence intervals
cit_rep <- confint(regx_rep, level = 0.9)
## plot LM statistics along with confidence interval
flm_rep <- fxlm(CNY ~ USD + JPY + EUR + GBP, data = cny_rep)
scus_rep <- gefp(flm_rep, fit = NULL)
plot(scus_rep, functional = supLM(0.1))
## add lines related to breaks to your plot
apply(cny_rep,1,function(x) sum(is.na(x)))

We got breaks in 2010 and in 2015 (when China’s stock markets crashed). We would have hoped for more breaks (we can still get them), but that would depend on the parameters chosen for our regression.