Which of the following is a macroeconomic factor that can affect a firm strategy?

13.8 Concluding remarks

This chapter compares different specifications with macroeconomic factors by allowing for threshold GAPMs driven by interest rate movements. From the structural break results, we see that the differing exposures to interest rate movements are not captured by a heterogeneous beta model, but by a double alpha effect for low beta portfolios. However, this method fails to find the impact of actual interest rate changes on the slope and intercept of the two models when there are different changes in the same period.

In our proposed specification, using the sign of the interest rate change (validated by a reference point check using a grid search upon the likelihood function of our specification) rather than the actual change, we find that alpha is negative for low beta portfolios whenever the interest rate is rising and that it is positive whenever the rate is decreasing. In line with the previous results, we find significant evidence of outperformance of low beta portfolios based on interest rate movements and underperformance of high beta portfolios. There is no systematic effect of the interest rate on beta itself. This is evidence that the outperformance of low beta portfolios is not related to their systematic market risk but to interest rate factors that influence the intercept of the GAPM.

We show that the opaque nature of the definition of the riskless asset is a complicating factor. We find evidence that the slope of the yield curve has a significant and differentiating impact on low and high beta portfolios by using a simple general equilibrium model. We consider 1 month, 10-year rates and an equilibrium combination of the two based on an estimated relative share of investors. We might expect that the appropriate rate for the GAPM is the 1-month rate as this would reflect the rebalancing period of institutional investors. What we find empirically is that we see similar results for the slope of the yield curve and the long-term rate.

When we test a misspecified version of the GAPM based on a mismatch in maturity levels and investor preferences, we observe that the short-term interest rate does not have a significant impact on the excess returns of the portfolios, in line with theory. However, we expect the sign of the long-term rate to be positive in both cases. We find that the coefficient for the low beta portfolio is of the opposite sign, resulting in a rejection of the hypothesis that the anomaly arises from this particular form of mismeasurement. However, the analysis might differ if we include more securities of different maturities.

The main force behind the anomaly is likely to be attributed to exogenous macroeconomic factors influencing the risk-free rate. Monetary policy over the last 30 years has favored low beta strategies by increasing the price of bonds and it is fair to say that these macroeconomic factors shape our results, and are the main drivers behind off-equilibrium movements of returns. Hence, our model provides a link between macroeconomic (yield curve related) factors and the origin of the low beta anomaly. It seems that the underlying exposure to the risk-free asset has to be considered for a model consistent with the CAPM implications. To call out of equilibrium movements, an anomaly in the social sciences seems unwarranted.

3.5.4 Conditional Regression Models

Any of the above regression models can be augmented with macroeconomic factors, if the econometrician (investor) believes that alpha or beta is time-varying and varies with macroeconomic cycles. While we explore this topic in depth in a Chapter 7, we present some introductory models here.

First, suppose that we believe that asset managers have betas that are a function of one or more macroeconomic variables, such as interest rates, inflation, default spreads, industrial production, consumer confidence, the slope of the term structure of government interest rates, etc. Suppose that we are evaluating U.S. equity managers, and wish to use the four-factor model, augmented with macro factors. Here, we could use a conditional version of the four-factor model that controls for time-varying RMRFt loadings by a mutual fund, using the technique of Ferson and Schadt (1996) as follows:

(3.10)ri,t =αi+βi·RMRFt+si·SMBt+hi·HMLt+ui·UMDt+∑j=1KBi,j[zj,t-1·RMRFt]+εi,t

In this model, there are K macro-factors (which must be carefully chosen by the econometrician), and the levels of the macro factors are measured at the end of the prior month, t-1. Note that we have assumed that the macro-factors only affect the level of the loading on RMRF, but we could add similar macroeconomic interaction variables for the other three risk factors. For example, we can add ∑j=1KBi,j[zj,t-1·SMBt] if we think that the small stock premium varies with the macroeconomic cycle.

If we, in addition, believe that manager skills vary with macroeconomic conditions, we can follow the Christopherson, Ferson and Glassman (1998) conditional framework that allows both the alpha and the factor loadings of a fund to vary through time. For example, the above model is modified as follows:

ri,t=αi+∑j=1KAi,j·zj,t-1+β i·RMRFt++si·SMBt +hi·HMLt+ui·UMDt+∑j=1KBi,j[zj,t-1·RMRFt ]+εi,t,

This model computes the alpha of a managed portfolio, controlling for any investment strategies that use publicly available economic information to change either the portfolio’s beta or the portfolio’s allocation to individual stocks with abnormally high expected returns, conditional on the information. In this model, the skill of a manager is measured as αi+∑j=1K Ai,j·zj,t-1. The first term, αi, measures the baseline skill of the manager over all types of business conditions, while the second term, ∑j=1K Ai,j·zj,t-1, measures the additional skill (which might be negative or positive) due to time-varying skill effects.

Macroeconomic Factor Models

A macroeconomic multi-factor model defines a relationship between stock returns and a set of macroeconomic variables such as GDP, inflation, industrial production, bond yields, etc. The appeal of using macroeconomic data as the explanatory factors in the returns model is that these variables are readily measurable and have real economic meaning.

While macroeconomic models offer key insight into the general state of the economy they may not sufficiently capture the most accurate correlation structure of price movement across stocks. Additionally, macroeconomic models may not do a good job capturing the covariance of price movement across stocks in “new economies” or a “shifting regime” such as the sudden arrival of the financial crisis beginning in Sep. 2008.

Ross, Roll, and Chen (1986) identified the following four macroeconomic factors as having significant explanatory power with stock return. These strong relationships still hold today and are:

Unanticipated changes in inflation.

Unanticipated changes in industrial production.

Unanticipated changes in the yield between high-grade and low-grade corporate bonds.

Unanticipated changes in the yield between long-term government bonds and t-bills. This is the slope of the term structure.

Other macroeconomic factors have also been incorporated into these models include change in interest rates, growth rates, GDP, capital investment, unemployment, oil prices, housing starts, exchange rates, etc. The parameters are determined via regression analysis using monthly data over a five-year period, e.g., 60 observations.

It is often assumed that the macroeconomic factors used in the model are uncorrelated and analysts do not make any adjustment for correlation across returns. But improvements can be made to the model following the adjustment process described above.

A k-factor macroeconomic model has the form:

(6.40)ri= αi0+bˆi1f1+ bˆi2f2+⋯+bˆik fk+ei

Analysts need to estimate the risk exposures bik s to these macro-economic factors.

Cross-Sectional Multi-Factor Models

Cross-sectional models estimate stock returns from a set of variables that are specific to each company rather than through factors that are common across all stocks. Cross-sectional models use stock specific factors that are based on fundamental and technical data. The fundamental data consists of company characteristics and balance sheet information. The technical data (also called market driven) consists of trading activity metrics such as average daily trading volume, price momentum, size, etc.

Because of the reliance on fundamental data, many authors use the term “fundamental model” instead of cross-sectional model. The rationale behind the cross-sectional model is similar to the rationale behind the macroeconomic model. Since managers and decision-makers incorporate fundamental and technical analysis into their stock selection process it is only reasonable that these factors provide insight into return and risk for those stocks. Otherwise why would they be used?

Fama and French (1992) found that three factors consisting of (1) market returns, (2) company size (market capitalization), and (3) book to market ratio have considerable explanatory power. While the exact measure of these variables remains a topic of much discussion in academia, notice that the last two factors in the Fama-French model are company specific fundamental data.

While many may find it intuitive to incorporate cross-sectional data into multi-factor models, these models have some limitations. First, data requirements are cumbersome requiring analysts to develop models using company specific data (each company has its own set of factors). Second, it is often difficult to find a consistent set of robust factors across stocks that provide strong explanatory power. Ross and Roll had difficulty determining a set of factors that provided more explanatory power than the macroeconomic models without introducing excessive multicollinearity into the data (Figure 6.9).

Figure 6.9. Eigenvalue-Eigenvector Decomposition of a 100 Stock Portfolio

The cross-sectional model is derived from company specific variables, referred to as company factor loadings. The parameters are typically determined via regression analysis using monthly data over a longer period of time, e.g., a five-year period, with 60 monthly observations.

The cross-sectional model is written as:

(6.41)rit= xi1*fˆ1t+xi2 *fˆ2t+⋯+xik*fˆkt+eit

where xij* is the normalized factor loading of company i to factor j. For example,

xkl*=xkl−E(xk)σ(xk)

where E(xk) is the mean of xk across all stocks and σ(xk ) is the standard deviation of xk across all stocks.

And unlike the previous models where the factors were known in advance and we estimate the risk sensitivities, here we know the factor loadings (from company data) and we need to estimate the factors.

4.5 Factors as Fundamentals

In our discussion of the links between market risk and macro fundamentals we have sometimes been casual in distinguishing returns from excess returns, realized from expected returns, realized from expected volatility, and related, in our treatment of timing. This is to some extent unavoidable, reflecting different conventions both within and among different and evolving literatures, as well as our desire to convey wide-ranging ideas in this broad survey. Nevertheless, a clearly emergent theme is that financial markets, as summarized by μr and σr, are very much linked to the business cycle, as summarized by μf and σf. Indeed it is not an exaggeration to claim that business-cycle risk may be the key driver of expected excess equity returns and return volatilities. Here we expand on that insight.

Although the business cycle may be a key risk factor, a long tradition, dating at least to Burns and Mitchell (1946) and actively extending to the present, recognizes that no single observed variable is “the business cycle” or “real activity”. Instead, we observe literally dozens of indicators (employment, industrial production, GDP, personal income, etc.), all of which contain information about the business cycle, which is not directly observable. Hence the key business cycle real activity fundamental underlying risk may be appropriately and productively viewed as a common factor to be extracted from many individual real activity indicators.

Expanding on this “factors as fundamentals” perspective, another likely relevant additional factor candidate is price/wage pressure, which may of course interact with real activity, as emphasized in Aruoba and Diebold (2010). In any event, the point is simply that, although we see hundreds of macroeconomic fundamentals, a drastically smaller set of underlying macroeconomic factors is likely relevant for tracking market risk. This is useful not only for best-practice firm-level risk management, but also for regulators. In particular, the factors-as-fundamentals perspective has important implications for the design of stress tests that simulate financial market responses to fundamental shocks, suggesting that only a few key fundamentals (factors) need be stressed.

Not surprisingly, then, we advocate that risk managers pay closer attention to macroeconomic factors, as they are the ultimate drivers of market risk. We hasten to add, however, that due to the frequent “disconnect” problems mentioned earlier, we would never advocate conditioning risk assessments only on macroeconomic factors. Rather, macroeconomic factors complement, rather than substitute, for the methods discussed in earlier sections, by broadening the conditioning information set to include fundamentals in addition to past returns.One might reasonably question the usefulness of conditioning on macroeconomic data for daily risk assessment, because macroeconomic data are typically available only quarterly (e.g. GDP and its components), or sometimes monthly (e.g. industrial production and the CPI). Recent developments that exploit state space methods and optimal filtering, however, facilitate high-frequency (e.g. daily) monitoring of latent macroeconomic fundamental factors. In particular, based on the high-frequency real activity monitoring approach of Aruoba, Diebold, and Scotti (2009), the Federal Reserve Bank of Philadelphia produces the “ADS index” of real activity, updated and written to the web in real time as new indicator data, released at different frequencies, are released or revised.96

We have emphasized macroeconomic fundamentals for equity market risk, but the bond market is also closely linked to macroeconomic fundamentals. In particular, government bond yield curves are driven by just a few factors (level, slope, curvature), with the level factor closely linked to price/wage activity and the slope factor closely linked to real activity.97 The same is true for yield curves of defaultable bonds, except that there is the additional complication of default risk, but that too is linked to the business cycle. Hence despite data on dozens of government bond yields, and dozens of macroeconomic indicators, the interesting reality is their much lower-dimensional “state vectors”—the level and slope factors beneath the yield curve, and the real and price/wage activity factors beneath the macroeconomy. One can easily imagine the usefulness for daily market and credit risk management (say) of systems linking yield curve factors (level, slope, curvature, …), equity factors (market, HML, SMB, momentum, liquidity, …), and macroeconomic factors (real, price/wage, …). All of those factors are now readily available at daily frequency.

4.2.2 Cross-Country Incidence

The substantial increase in countries adopting AD laws in the 1980s and 1990s, along with the establishment of the WTO, which began collecting data more systematically from member countries, led to a literature examining the incidence of AD activity across countries. Miranda et al. (1998), Prusa (2001, 2005), and Zanardi (2004a) document and assess the proliferation of countries with AD laws and resulting AD activity. Bown (2010, 2011a,b) discusses more recent patterns of AD usage across countries.

Knetter and Prusa (2002) examine aggregate AD activity over time and focus on macroeconomic factors.z Knetter and Prusa demonstrate that a depreciation of the exporters’ exchange rate will have conflicting effects on the dumping margin and injury test. A depreciation will lower the exporters’ price which therefore increases the prospect of injury. On the other hand, given that pass through is incomplete, the change in pricing will decrease the prospect of less than fair value sales. In theory, either effect could dominate. In practice, they find that periods of poor GDP growth and strong currency are positively correlated with increased country-level AD activity. In other words, macroeconomic forces (ie, business cycles) and exchange rate movements affect filings of AD petitions and the likelihood of successful AD decisions.aa We note that the allegedly dumping exporting firms have no say in where a country is in the business cycle or the value of the exchange rate. Consequently, one might consider how much AD activity is really about economically meaningful dumping vs the need to provide protection to politically important industries.

Other studies examine strategic interdependence of AD activity across countries. Maur (1998) documents the correlation in the industry usage of AD across countries, a phenomenon he calls “echoing.” The steel crisis during the late 1990s/early 2000s is a stark example of echoing cases (Durling and Prusa, 2006).

Prusa and Skeath (2002, 2005) and Feinberg and Reynolds (2006) find evidence that countries may be engaging in tit-for-tat AD actions against each other. In effect, countries appear to be using the flexibility of antidumping to raise the cost of partners using AD. While the WTO does not allow for compensation for AD duties, countries appear to be able to use their own AD as unofficial retaliation.

In contrast, Blonigen and Bown (2003) find that US petitions for AD duties against foreign firms in another country are less likely when the US industry has significant exports to that same foreign country. The study also finds that the US government is less likely to rule favorably on a petition for an AD duty when the named country is a WTO member who can retaliate by filing a request for dispute settlement with the WTO Dispute Settlement Body.

Publisher Summary

This chapter examines labor relations during the regime of the former President, Kim Young Sam, emphasizing macroeconomic, political, and institutional factors taking into consideration that the Korean labor unions are decentralized. The chapter provides an overview over the Korean economic development by presidential regimes. It compares the misery and economic performance indices by presidential regimes. Two turning points in the Korean industrial relations have been recognized. The first turning point occurred in the June 29, 1987 democratization declaration by President Roh Tae Woo, which marks a rise in wages, strikes, and labor market institutions. The second turning point occurred in the early 1990s, which marks a decline in wages, strikes, and labor market institutions without full-fledged industrial reforms. The chapter reviews Korean industrial relations by the presidential regimes from 1945 to 1990. It discusses the industrial relations during the Kim Young Sam regime, which includes an extensive examination of the wage policy and wage-price-productivity nexus.

Learn From Reality

What is the future of libraries in this world of rapid change? Society transforms over time and is shaped by the macroeconomic factors, technological shifts and peoples’ behaviour. Libraries do not exist in isolation. In addition to their historical roles and missions, they need to transform and move along with current trends to avoid becoming obsolete. Libraries are key components of a knowledge society, supporting patrons to satisfy their variety of needs. Many learning activities could be conducted in facilities other than libraries. Why do we need libraries? It is useful to observe and learn from recent history, especially from the business world. Kodak and Polaroid faced enormous challenges as film was replaced by digital cameras. Digital cameras also face the risk of being replaced by smartphones. Only businesses that are adaptable and can co-evolve with new and disruptive technology will stay competitive. This also applies to academic libraries.

Customer Needs and Preferences

It is important to understand the customer needs and preferences in the market to know what investors and operators are willing to pay for an aircraft. Macroeconomic factors should not be the sole determining factor in the decision to develop and produce an aircraft as it can sometimes lead to wrong decisions. For example, fuel prices rose drastically in the late 1990s through 2008, and airlines and lessors demanded more fuel-efficient aircraft as well as larger aircraft to reduce frequency and CASM. Airbus and Boeing thought it best to focus primarily on fuel efficient aircraft such as the A320neo and the B737 MAX, as well as large aircraft such as the A380 due to the facts at the time. After fuel prices dropped, airlines found they could continue to operate their older, less fuel-efficient aircraft economically, and demand for new, more fuel efficient aircraft softened. In addition, after the introduction of the A380, it was discovered that there are limited markets for an aircraft of this size, as passengers prefer higher frequencies on smaller aircraft than the lower frequency of a larger aircraft.

Lessors and investors will focus on asset liquidity in connection with yield. Liquidity is almost synonymous with risk. If the aircraft is very liquid, risk is much lower, which therefore means the investor may be satisfied with a lower return. If a lessor is seeking higher returns, they should invest in less liquid assets to receive a premium. For example, the B737-800 is considered a highly liquid aircraft. However, the returns on this aircraft for a lessor tend to be lower due to large numbers of financiers who are offering creditworthy airlines low-interest rates to finance the aircraft, forcing lessors to offer low lease rates to compete. Widebodies such as the A380 or B777 have a much narrower customer base, making it much more difficult to redeploy the airplane and lessors face additional risk resulting from the costs associated with reconfiguration. The number of routes and operating customers is smaller, so the risks are much higher and few parties are willing to invest, generating seemingly higher returns assuming the investor’s residual value assumption plays out as expected. When examining today’s investment profile, almost half of the narrowbody fleet is owned by investors rather than the airlines, whereas only 30% of widebodies are owned by financial investors, and the remaining 70% are owned by airlines due to the higher risk associated with widebodies and the secondary market.

Another example of a category of aircraft which customers have expressed an interest in is what is referred to as the “middle of the market” aircraft. The gap between the A321 and the B787-8 in terms of size is wide enough to warrant an aircraft similar to, or slightly larger than, the B757. Airlines have the requirement, but manufacturers are hesitant to produce it because of the significant development cost associated with designing and producing a new aircraft type. Thus, manufacturers are not willing to commit to a new airframe program without a solid economic case, and that means enough demand to far exceed the development costs, leading to a profitable program.

Aircraft finance

These financings of individual aircraft tend to be long term, commonly at least 12 years, and are therefore seen as a very specialized sector. They combine macro-economic factors such as the number of passenger or freight flights per annum, together with micro-economic factors such as the fixture stability of an airline, the statutory periodic airworthiness certificates required on individual aircraft and the likelihood of fixture changes to regulations for factors such as noise pollution, and the future valuation and/or usefulness of an airplane if it is no longer required by the airline which commissioned its construction.

Aircraft finance loans are regularly seen in the syndicated loans market but are generally confined to participants who specialize in the sector.

6.4.3. The Cohort Approach

Let us now consider the retail loans and introduce a dynamic model for data on default rates aggregated by cohorts, which is easy to estimate and simulate. This model includes some autoregressive effects of lagged default and macroeconomic factors and accounts for unobserved time heterogeneity. Because of the autoregressive component, the specification for the first semester of a loan agreement of any term is different from the specification for the next semester and the following ones. In the first semester of the loan, there is no information on past default history of the cohort. The initial credit quality of that cohort can be approximated by using the basic credit score of the credit granting institution. Let Sk,τ and σ2S,k,τ denote the average initial score and its dispersion in cohort k, τ. In semester h = 1, we use the following logistic model:

(6.11)l [Dk(τ;1)]=a1+b1l(Sk,τ)+c1σS,k,τ2+d′1Xτ+1+α1l[Dk(τ -1;1)]+ɛk(τ;1),

where the components of X are macroeconomic variables, εk(τ; 1) is an error term, and l(x) = log[x/(1 — x)] denotes the logit transformation. For the following semesters of the loan, we introduce an additional autoregressive effect of the same cohort (τ, k) and a lagged effect of the previous cohort (τ — 1, k):

(6.12)l[Dk(τ ;h)]=ah+bhl(sk,τ) +chσs,k,τ2+d′hXτ+h+αhl[Dk(τ-1;h)]+βhl[Dk(τ,h-1)]+ɛk( τ,h),h≥2.

The joint model [(6.11) and (6.12)] is a spatial regression model. It is completed byspecifying the distribution of error term εk(τ, h), for any k, τ, h. Let us assume the independence between cohorts and allow for correlation between semesters. More precisely, we assume

[εk(τ, h), h = 1,…, H], τ, k varying, are independent, normally distributed, with mean zero and variance-covariance matrix Σ.

Parameters ah, bh,…,βh, h= 1,…, H and Σ can be estimated by the ordinary least squares. Even though the number of parameters is large, in each semester h, the number of available observations is too large and equal to the number of cohorts times the number of loans with different terms.

The estimated models can be used for prediction making. In particular, the following columns of Table 10.3 can be found by simulations. For example, for the future date 00.1 (first calendar semester of year 2000), error ∈s(00.1; 1) is drawn and the simulated default rate Ds(00.1; 1) is determined by model (6.11) from D(99.2; 1) and the simulated error. For the second row of that column, we simulate error ∈s (99.2; 2) and use Ds(99.2; 2) determined by model (6.12), D(99.2; 1), D(99.1; 1) and ∈s(99.2; 2), and so forth. Note the difficulty in predicting the future values of macroeconomic variables X. A solution consists in considering several scenarios of their future evolution to assess the default rate.