Market timing, factors and sectors
Rob Hayward
Introduction
Recall that Modern Portfolio Theory (MPT) says that investor should hold the market portfolio, all the assets in the investment universe weighted according to their capitalisation to achieve the maximum return per unit of risk. The level of risk can be determined by mixing the market portfolio with the risk-free asset. This is equivalent to optimising the efficient frontier on the mean, variance and covariance of the assets in the investment universe.
The investment universe can be defined in different ways and there might be an argument to change the level of risk according to market conditions.
Market timing
Market timing is a style of investment that will seek to switch between the risk-free asset and the market portfolio depending on market conditions. Practically, this can mean switching between stocks and bonds or between a basket of assets and cash. It be developed so that more complex conditions are applied to determine the switching conditions. Most research suggests that market-timing, like security-selection, is unlikely to be successful.
Practice 3.1
- Use the data that we downloaded last week (BACSPY.csv) to asses the times when it would have been helpful to have been out of the market. You need some indicators that would have forewarned you about these equity difficulties. What could those be?
Market timing can also relate to:
Equities: industry sectors. Investors may try to time the movement into cyclical stocks to coincide with an upturn in the economy.
Fixed income: duration and credit. Investors will take more duration and credit risk according to their assessment of economic and financial conditions. For example, longer-maturity bonds will be more affected by interest rate changes; corporate bonds will suffer from defaults during a recession but government bonds will be supported by flight-to-quality.
Commodities: cyclical and structural. Investors may target cyclical commodities (steel, copper etc), structural changes that may affect supply and demand (i.e. climate change may reduce the demand for crude oil while raising the demand for trace metals that are used in batteries). Gold and silver are often used for protection against inflation or global economic and financial uncertainty.
Factor investment
Recall CAPM identifies the following equation:
\[ R_i = \alpha_i + \beta_i (MR -rf)\]
Where R is the return for security i, MR is the market return, rf is the risk-free rate, beta is the amount of market risk in the individual security and alpha is the excess return on top of the return for taking risk (beta). Market efficiency says that there is no alpha.
It would be possible to extend this analysis to include other risk factors and assess whether there is alpha once risk factors beyond the standard market risk have been removed. This is the sort of factor-based advanced risk management technique that is used by hedge funds.
Beta is the market factor or equity risk premium. However, pretty soon market participants and academics found that beta was not very good at describing the returns for individual securities. Other factors have been identified. These include value and small capitalisation. Growth is usually regarded as a negative factor (though that can be debated). A factor zoo has been added, including momentum, quality, defensive. Take a look at the FTW function on Bloomberg. Term and risk factors have been identified for bonds.
Fama and French identified a three factor model. Eugene F. Fama and French (1993) and followed that up with a five factor model. Eugene F. Fama and French (2015). The factors that they identified.
Equity risk premium. This is the beta.
Size. Market capitalisation. Also called SMB as it may be created by buying small capitalisation stocks and selling large capitalisation stocks.
Value. Cheap relative to fundamentals. Also called HML as it may be created by buying companies with a high book-to-market valuation relative to those with low book-to-market valuation.
Quality. Robust earnings. Also called RMW as it may be created by buying companies with high or robust operating profitability and selling those with low or weak operating profitability.
Defensive. Cautious investment. Also called CMA as it may be created by buying companies with conservative investment activity relative to those with aggressive investment.
Factor models
There are three main ways to look at factor performance.
Identify individual stocks that have high factor values. This is the Buffet approach. It is called value investment as he is said to seek investment in firms that are relatively cheap compared to the fundamentals.
Build a portfolio of stocks with high factor values that is short low factor values and compare that to average stock returns. This is the hedge fund approach (more next week).
Find exchange traded funds (ETF) that are based on factors. There are many funds, but they are more expensive than simple index funds because there must be rules about selection and rebalancing and there will be trading costs.
Practice 3.2
Go to the Ken French factor library and download the current daily data for the standard Fama/French 5 Factors (2x3) (Daily) model. Read the Details to understand the data.
Import the data into R, ensure that the date is a date object.
Plot the 5 factors performance since July 1963. Which is the best performing? What are the cycles of performance?
Growth and value
The two core factors are growth and value. These can also be seen as representing optimism about future profits (growth) or defensive demand for strong fundamentals. Though the Buffet strategy and the original Fama-French papers suggest that value will outperform, this has not been the case in recent years. US technology (growth) stocks have been leading the way.
It is possible to get a view of evolving sentiment about growth (optimism) relative to value (defensive) views by charting ETFs for growth and value factors.
require(quantmod)
# Growth VUG and Value VTV
mysymbols <- c('VUG', 'VTV')
getSymbols(mysymbols)## [1] "VUG" "VTV"mydata <- merge(Cl(VUG), Cl(VTV))
da <- as.data.frame(mydata)
plot(as.Date(rownames(da)), da$VUG.Close/da$VTV.Close, type = 'l',
main = 'Growth and Value indices', xlab = 'Date', ylab = 'Ratio')
A move higher means that growth is out-performing, which is the case for most of these eight years. However, you can see a very positive period for value in March 2020 (Covid), in 2023 (inflation) and in the most recent data.
Extending the factor model
This basic analysis can be extended to other asset classes and it can be adapted to different industries. Some examples:
- Fixed income factors include
- term premium
- credit
- volatility
- Commodities are usually split between
- energy (oil and gas)
- softs (agricultural products or livestock, such as wheat, corn,
coffee etc)
- metals (precious such as gold, sliver and platinum; base metals such as copper, aluminium and zinc)
| Sector | ETF |
|---|---|
| Information technology | XLK |
| Consumer discretionary | XLY |
| Utilities | XLU |
| Financials | XLF |
| Communication services | XLC |
| Real estate | XLRE |
| Healthcare | XLV |
| Industrials | XLI |
| Consumer staples | XLP |
| Materials | XLB |
| Energy | XLE |
These sectors may change with the economic cycle. For example, cyclical sectors are usually considered to be consumer discretionary, real estate, financials and materials on the basis that households will spend more on expensive goods and that there will be fewer bad loans and more demand for raw materials as the economy expands; healthcare, consumer staples and utilities are considered to defensives that will outperform when the economy is doing badly.
Practice 3.4
Download a sector index and assess the performance during periods of positive and negative economic growth.
Does it perform as theory would suggest?