Capital Market Expectations

• A Framework for Developing Capital Market Expectations
• Challenges in forecasting
  • Limitations of Economic Data
  • Data Measurement Errors and Biases
  • The Limitations of Historical Estimates
  • Biases and Uncertainties in Analysts’ Methods
  • Shrinkage Estimate
• Economics Growth
  • Exogenous Shocks to Growth
  • Trend Growth after a Financial Crisis
  • Decomposition of GDP Growth
• Approaches to Economic Forecasting
  • Econometric Modeling
    • Advantages
    • Disadvantages
  • Economic Indicators
    • Advantages
    • Disadvantages
  • Checklist Approach
    • Advantages
    • Disadvantages
• Business Cycle
  • Initial Recovery
  • Early Expansion
  • Late Expansion
  • Slowdown
  • Contraction
• Inflation Expectation
  • Inflation at or below expectations
  • Inflation above expectations
  • Deflation
• Fiscal and Monetary Policy
  • Taylor’s Rule
• International Interactions
  • The Impossible Trinity
• Approaches to Forecast Asset Values
  • The Building Block Approach to Fixed-Income Returns
  • DCF Approach to Equity Returns
    • Grinold–Kroner model
  • Risk Premium Approaches to Equity Returns
    • Singer and Terhaar model
• Forecasting Real Estate Return
• Forecasting Exchange Rates
  • Purchase Power Parity
  • Uncovered Interest Rate Partiy
• Forecasting Volatilities
  • Sample Statistics for CVC Matrix
    • Advantages
    • Disadvantages
  • Factor Model for CVC Matrix
    • Advantages
    • Disadvantages
  • Shrinkage Estimation for CVC Matrix
  • Estimating Volatility from Smoothed Returns
  • Time-Varying Volatility: ARCH Models

A Framework for Developing Capital Market Expectations

1. Specify the set of expectations needed, including the time horizon(s) to which they apply.
2. Research the historical record.
3. Specify the method(s) and/or model(s) to be used and their information requirements.
4. Determine the best sources for information needs.
5. Interpret the current investment environment using the selected data and methods, applying experience and judgment.
6. Provide the set of expectations needed, documenting conclusions.
7. Monitor actual outcomes and compare them with expectations, providing feedback to improve the expectations-setting process.

Challenges in forecasting

Limitations of Economic Data

• Time lag
• Revisions to data
• Changes in definition and methodologies

Data Measurement Errors and Biases

• Transcription errors
• Survivorship bias
• Appraisal (smoothed) data. For certain assets without liquid public markets, appraisal data are used in lieu of transaction data. Appraised values tend to be less volatile than market-determined values. As a result, measured volatilities are biased downward and correlations with other assets tend to be understated.

The Limitations of Historical Estimates

• Regime change: Changes that stem from technological, political, legal, and regulatory environments.
• Regime changes give rise to the statistical problem of nonstationarity

Biases and Uncertainties in Analysts’ Methods

• Data-mining bias: repeatedly searching a dataset until a statistically significant pattern emerges. Lack of an explicit economic rationale for a variable’s usefulness is a warning sign of a data-mining problem:
• Time-period bias ：Research findings often turn out to be sensitive to the selection of specific starting and/or ending dates
• Model uncertainty: whether a selected model is structurally and/or conceptually correct.
• Parameter uncertainty: parameters are invariably estimated with error.
• Input uncertainty: inputs could be incorrect or the need to proxy for an unobservable variable

Shrinkage Estimate

Shrinkage estimation involves taking a weighted average of a historical estimate of a parameter and some other parameter estimate

Economics Growth

Exogenous Shocks to Growth

• Policy changes
• New products and technologies
• Geopolitics
• Natural disasters
• Natural resources/critical inputs.
• Financial crises

Trend Growth after a Financial Crisis

• Type 1: A persistent (permanent, one-time) decline in the level of output, but the subsequent trend rate of growth is unchanged.
• Type 2: No persistent decline in the level of output, but the subsequent trend rate of growth is reduced.
• Type 3: Both a persistent decline in the level of output and a reduction in the subsequent trend rate of growth.

Decomposition of GDP Growth

• growth from labor inputs, consisting of
  • growth in potential labor force size and
  • growth in actual labor force participation
• growth from labor productivity, consisting of
  • growth from increasing capital inputs (capital deepening)
  • growth in total factor productivity

Aggregate market value of equity, Ve = Nominal GDP * the share of profits in the economy, Sk (earnings/GDP) * the P/E ratio (PE).

Approaches to Economic Forecasting

Econometric Modeling

• Econometrics: application of quantitative modeling and analysis grounded in economic theory to the analysis of economic data.
• Structural models: specify functional relationships among variables based on economic theory.
• Reduced-form model: more compact representations of underlying structural models. Evaluate endogenous variables in terms of observable exogenous variables.

Advantages

• Robust models
• New data may be collected and consistently used within models to quickly generate output.
• Delivers quantitative forecasts
• Imposes discipline/consistency on analysis.

Disadvantages

• Complex and time-consuming to formulate.
• Relationships not static. Model may be mis-specified.
• May give false sense of precision
• Rarely forecasts turning points well

Economic Indicators

• Leading economic indicator
• Lagging economic indicator
• Diffusion Index: an index that measures how many indicators are pointing up and how many are pointing down.
• Look-Ahead Bias: a study relies on data or information that is not yet available or known during the study

Advantages

• Usually intuitive and simple in construction.
• Focuses primarily on identifying turning points.
• Data available from third parties. Easy to track.

Disadvantages

• History subject to frequent revision
• Overfitted in sample. Likely overstates forecast accuracy.
• “Current” data not reliable as input for historical analysis.
• Can provide false signals.
• May provide little more than binary (no/yes) directional guidance.

Checklist Approach

• Consider a whole range of economic data to assess the economy’s future position.

Advantages

• Flexible. Structural changes easily incorporated. Items easily added/dropped.
• Breadth: Can include virtually any topics, perspectives, theories, and assumptions.

Disadvantages

• Subjective. Arbitrary. Judgmental.
• Time-consuming. Manual process limits depth of analysis.
• No clear mechanism for combining disparate information.
• Imposes no consistency of analysis across items or at different points in time.

Business Cycle

Initial Recovery

• The economy picks up, business confidence rises, an upturn in spending on housing and consumer durables.
• Stimulative policies are still in place. Transitioning to tightening mode.
• Negative output gap is large
• Inflation is typically decelerating
• Rates are low and bottoming.
• Yield curve is steep. Shortest yields begin to rise first.
• Cyclical assets and riskier assets typically perform well.

Early Expansion

• Unemployment starts to fall
• Output gap remains negative
• Businesses step up production and investment
• Profits typically rise rapidly.
• Short rates are moving up as the central bank starts to withdraw stimulus
• Yields rising. The yield curve is flattening.

Late Expansion

• Output gap closes
• Unemployment is low, profits are strong
• Both wages and inflation are rising
• Debt coverage ratios may deteriorate as balance sheets expand and interest rates rise.
• Interest rates are typically rising as monetary policy becomes restrictive.
• Stock markets often rise but may be volatile as nervous investors endeavor to detect signs of looming deceleration.
• Yield curve continues to flatten
• Cyclical assets may underperform while inflation hedges such as commodities outperform.

Slowdown

• Economy is slowing and approaching the eventual peak
• Inflation often continues to rise
• Monetary policies become restrictive.
• The yield curve may invert. Credit spreads generally widen.
• The stock market may fall, with interest-sensitive stocks such as utilities underperforming and “quality” stocks with stable earnings performing best.

Contraction

• Firms cut production sharply. Profits drop sharply.
• Unemployment can rise quickly
• Monetary policies more stimulating to combat downward momentum
• Yields declining.
• The yield curve steepens substantially.
• The stock market declines in the earlier stages of the contraction but usually starts to rise in the later stages, well before the recovery emerges.
• Credit spreads typically widen and remain elevated until signs of a trough emerge

Inflation Expectation

Inflation at or below expectations

• Cash Equivalents (CE) and Bonds: Neutral with stable or declining yields
• Bonds: Positive as rates decreases and prices increases. Persistent deflation benefits the highest-quality bonds because it increases the purchasing power of their cash flows. It will, however, impair the creditworthiness of lower-quality debt.
• Equity: Positive with predictable economic growth
• Real Estate (RE): Neutral with typical rates of return

Inflation above expectations

• CE: Positive with increasing yields
• Bonds: Negative as rates increase and prices decline
• Equity: Negative, though some companies may be able to pass through inflation and do well
• RE: Positive as real asset values increase with inflation

Deflation

• CE: Negative with approximately 0% interest rates
• Bonds: Benefits the highest-quality bonds because it increases the purchasing power of the cash flows, but it is likely to impair the creditworthiness of lower-quality debt.
• Equity: Negative as economic activity and business declines
• RE: Negative as property values generally decline

Fiscal and Monetary Policy

• Fiscal policies: taxation and spending

• Monetary policeis: policy rates and liquidity provision

• Loose (tight) fiscal policy = High (low) real rates
• Loose (tight) monetary policy = High (low) expected inflation

Taylor’s Rule

i*, target nominal policy rate = r-neutral, real policy rate that would be targeted if growth is expected to be at trend and inflation on target + π-e, expected inflation rate + 0.5 * (Y-e, expected real GDP growth rates - Y-trend, long-term trend in the GDP growth rate) + 0.5 * (π-e, expected inflation rate - π-target, targeted inflation rate)

International Interactions

• Current Account: net exports of goods and services, net investment income flows, and unilateral transfers.
• Capital Account:net investment flows for Foreign Direct Investment (FDI) and Portfolio Investment (PI) flows

Y = C + I + G + X - M

Y = C + S (private savings) + T(tax)

(X, expoert - M, import) = (S, savings - I, investment) + (T, tax - G, gov spending)

Current Account = (S-I) + (T-G)

The Impossible Trinity

• Unrestricted capital flows
• Fixed exchange rate
• Independent monetary policy

Approaches to Forecast Asset Values

• Formal tools
  • Statistical Methods
    • Sample statistics: sample means, variances, and correlations
    • Shrinkage estimation: taking a weighted average of two estimates of the same parameter—one based on historical sample data and the other based on some other source or information
    • Time series estimation: forecasting a variable on the basis of lagged values of the variable being forecast
  • Discounted Cash Flow
  • Risk Premium Models
    • An equilibrium model, i.e. CAPM
    • A factor model
    • Building blocks
• Surveys
• Judgement

The Building Block Approach to Fixed-Income Returns

• The Short-term Default-free Rate
• The Term Premium
• The Credit Premium
• The Liquidity Premium

DCF Approach to Equity Returns

Grinold–Kroner model

E(Re) ≈ D/P + (%ΔE−%ΔS) + %ΔP/E Expected equity return ≈ dividend yield + expected changes in earnings - expected changes in S/O + expected changes in PE ratio

Risk Premium Approaches to Equity Returns

Singer and Terhaar model

1. Under the assumption of full integration with the global market portofolio: Risk Premium = β * GIM Risk Premium = correlation * (SD of asset / SD of GIM ) * GIM Risk Premium = correlation * SD of asset * GIM Sharpe Ratio

2. Under the assumption of full segmentation of markets, β = 1 and each asset is their own market Risk Premium = 1 * Market Risk Premium = 1 * SD of asset * (Risk premium of the asset / SD of asset) = Asset Risk * Asset Sharpe Ratio

3. The weighted average of two component: RPi = φ * Risk Premium Under Integration Assumption +(1−φ) * Risk Premium Under Segmentation Assumption

Forecasting Real Estate Return

Cap Rate = Net Operating Income / Property Value

Expected Return = Cap Rate + NOI growth rate

Forecasting Exchange Rates

Purchase Power Parity

• %Δ Ex Rate = %Δ Country A Inflation Expectation - %Δ Country B Inflation Expectation

Uncovered Interest Rate Partiy

• %Δ Ex Rate = %Δ Country A Nominal Interest Rate - %Δ Country B Nominal Interest Rate
• Yet the profitability of carry trade - borrowing in low-rate currencies and lending in high-rate currencies—violates the UIP assumption

Forecasting Volatilities

Sample Statistics for CVC Matrix

Advantages

• Unbiased and correct on average
• Converges to the true CVC matrix as sample size gets larger

Disadvantages

• Can’t be used for large numbers of assets
• Substantial sampling error unless number of observations is at least 10 times number of assets
• No cross sectional consistency

Factor Model for CVC Matrix

Advantages

• Can be used when number of assets exceeds number of observations
• Reduce the number of unique parameters to be estimated
• Imposes cross-sectional structure
• Substantially reduce estimation error

Disadvantages

• Mis-specified factor model will be biased and in consistent and will no converge on average

Shrinkage Estimation for CVC Matrix

• Weighted average of the two methods
• Captures the benefits of both methods while increase the efficiency (reduce MSE) of the estimates

Estimating Volatility from Smoothed Returns

• Current observed return, Rt, is a weighted average of the current true return, rt, and the previous observed return: Rt = (1−λ)*rt + λR_t−1
• var(r) = (1+λ)/(1−λ) * var(R) > var(R)

Time-Varying Volatility: ARCH Models

• Current variance depends only on the variance in the previous period and the unexpected component of the current return
• σ_t^2 = γ + α x σ_t−1^2 + β x η_t^2 = γ + (α+β)x σ_t−1^2 + β x (η_t^2 − σ_2t−1^2)