More Related Content
Similar to Risk under Uncertainty and Price Movement (20)
Risk under Uncertainty and Price Movement
- 1. Risk under
Parameter Uncertainty &
Price Movement
FRAGILITY–LATENT RISKS & UNSTABLE HEDGES–MADE VISIBLE
ANISH R. SHAH, CFA ANISHRS@INVESTMENTGRADEMODELING.COM
FEB 2, 2021 NETHERLANDS (FROM BOSTON)
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 2. Two slide overview (1/2)
1. Risk decomposition is a tool to understand portfolio’s net risk
2. Net risk = what remains after hedges cancel
For example, in market neutral, longs – shorts = hedged market effect of 0 + other risks
3. Hedged risks are often bigger (before canceling) than the net
Imagine the amount of market effect in each side of long/short
4. Numbers are inexact
→ ‘hedged’ risks, in truth, aren’t and contribute to the net
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 3. Two slide overview (2/2)
5. How can numbers be off? Two ways
Parameter uncertainty from estimation error
Dynamic portfolio composition over the horizon of interest
Price movement alters position weights
Hedges hold for an instant, fluctuate as weights shift
6. What do you do?
Regard everything as uncertain
Work from estimates of mean and variance instead of fixed values
For parameters – center and error generated during inference
For weights – distribution of future portfolio weights over horizon
7. Risk decomposition reports expected value ± standard deviation
Surfaces latent fragility
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 4. Mostly conceptual presentation
See papers for math
Risk decomposition with uncertainty
◦ Section V of Shah, A. (2019). Uncertain risk parity. http://ssrn.com/abstract=3406321
With uncertainty and price movement
◦ <coming> Shah, A. (2021). Uncertain and dynamic risk contributions. http://ssrn.com/abstract=3774239
Criticism and questions: AnishRS@InvestmentGradeModeling.com
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 5. Risk decomposition
S&P500 volatility by factors
Factor Exposure Risk contribution
Beta 0.95 16.32
Mkt Cap 2.53 2.10
Price Volatility -0.84 1.04
Relative Strength 0.39 0.49
Book/Price -0.50 0.38
⁞
Earnings Variability 0.09 -0.05
20.73
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
S&P500 volatility by sector buckets
Sector Wt Risk contribution
Information Technology 32.4 7.62
Consumer Discretionary 14.0 3.23
Health Care 15.0 2.87
Financials 9.7 1.98
Consumer Staples 8.0 1.53
⁞
Communication Services 2.3 0.29
100 20.73
Breaks portfolio’s volatility (in variance or std dev) into sources
- 6. Risk contributions geometrically
Menchero and Davis(2011). Risk
contribution is exposure times
volatility times correlation
Risks = directions (like north-south,
east-west)
Standard deviation = length
An investment (travel leg) is a
combination of risks
Portfolio = sum of investments
Risk contribution = length in the
direction of the portfolio
Movement in other directions
nets to zero
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
ATL
PHX
DEN
BOS
- 7. How can this go wrong?
Risk contributions are calculated from
◦ Covariance between securities
◦ Portfolio weights
Ingredients are inexact
◦ Risks that appear neutralized aren’t to some degree
Moreover, the conventional view considers only the net
◦ The magnitude of components involved doesn’t matter
e.g., Long beta=1 stock, short another beta=1 and
Long beta=3 stock, short another beta=3 appear the same
◦ In reality, the second has much more error and instability, contains large hidden risks
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
← estimated with error
← fluctuate with prices
- 8. How things go wrong in pictures
Imagine two directions of risk – market and
chicken feed – and investing long-short
Beta hedged → no net market risk → no
market in risk contributions
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
market
cf long
short
net
What if beta – estimated, not known – isn’t
perfectly hedged?
◦ The large market component in each side,
mismatched, drives the risk of the net
market
cf long
short
net
Perceived-as-hedged directions vanish regardless of sensitivity to underlying offsetting risks
◦ Imagine looking at the equity of a levered company without separately seeing the debt
◦ Hidden unstable risks at both the individual investment and portfolio levels
Optimization finds perceived hedges
market
cf long
short
net
- 9. Effect of price movement
Stability of hedges – hedged today doesn’t mean hedged tomorrow
Stylized example on two beta=1 portfolios:
Portfolio A: 50% beta=1 stock + 50% another beta=1 stock
◦ Portfolio maintains beta=1 regardless of price changes
Portfolio B: 50% cash + 50% beta=2 stock
◦ Market rises 10%. Portfolio is 45% cash + 55% stock, has beta=1.1
◦ Market falls 10%. Portfolio is 55% cash + 45% stock, has beta=0.9
◦ Beta exposure is constantly changing, unstable
Can quantify stability by assessing risk across the future distribution of portfolio weights
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 10. What do you do?
Problems
◦ Parameter estimates are unavoidably imperfect
◦ Portfolio composition changes with price movement
Solution
◦ Evaluate across the range of what’s possible
◦ Use parameters’ mean and covariance
◦ Model distribution of portfolio weights over the horizon of interest
◦ ± range accompanies numbers in risk decomposition
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 11. SP500 factor risk decomposition with
uncertainty but no price movement
Exposure ± Conventional Risk Uncertain Risk ±
Total 20.73 26.01 5.20
Beta 0.95 0.01 16.32 13.03 3.35
Mkt Cap 2.53 0.01 2.10 8.18 6.08
Price Volatility -0.84 0.12 1.04 2.43 2.45
Relative Strength 0.39 0.10 0.49 1.00 1.21
Book/Price -0.50 0.03 0.38 0.47 0.61
Trading Activity -0.26 0.06 0.13 0.31 0.42
⁞
Stock Specific 0.24 0.18 0.04
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
Note: uncertainty occurs in both exposures and factor/stock-specific variance levels
- 12. SP500 sector risk decomposition with
uncertainty but no price movement
Weight Conventional Risk Uncertain Risk ±
Total 100 20.73 26.01 5.20
Information Technology 32.4 7.62 10.05 2.27
Consumer Discretionary 14.0 3.23 3.99 0.84
Health Care 15.0 2.87 3.85 1.00
Financials 9.7 1.98 2.12 0.36
Consumer Staples 8.0 1.53 1.65 0.25
Industrials 7.3 1.17 1.71 0.51
Energy 2.9 0.68 0.61 0.19
⁞
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 13. Example of risk decomposition with
uncertainty and price movement
Total AWK TMUS CLX SBAC ODFL COG MKTX AMZN NEM …
Weight 100 20.0 18.2 15.3 14.6 10.5 9.8 6.4 1.8 1.3
Conventional Risk 19.5 1.64 1.60 1.35 1.29 0.93 0.87 0.57 0.16 0.12
5 day price
movement
Risk 19.6 1.62 1.59 1.34 1.28 0.91 0.86 0.56 0.16 0.12
± 1.9 0.17 0.15 0.11 0.19 0.13 0.17 0.09 0.02 0.01
Parameter
uncertainty
Risk 23.0 1.98 1.81 1.65 1.58 1.25 0.79 0.70 0.20 0.16
± 4.0 0.64 0.62 0.57 0.70 0.57 0.43 0.35 0.09 0.08
Both
Risk 23.2 1.96 1.79 1.63 1.56 1.24 0.77 0.69 0.20 0.16
± 4.7 0.72 0.69 0.61 0.78 0.63 0.53 0.38 0.09 0.08
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
Minimum variance long/short: 100% long, 100% short, 22 stocks – 1 in each sector on each side
- 14. Risk contributions – usual calculations
Portfolio = sum of items (defined according to how we want to cut risk)
◦ e.g., factor exposures, holdings by industry, holdings of individual securities
let σP
2 = var(portfolio), σk
2 = var(item k)
sk = standard deviation contribution from item k
= movement in direction of portfolio
= σk × corr(k, portfolio) = σk × cov(k, portfolio) / (σk σP)
= cov(k, portfolio) / σP
vk = variance contribution
= cov(k, portfolio)
Contributions properly sum to portfolio volatility: ∑k vk = σP
2 ∑k sk = σP
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
Differ only by σP
- 15. Risk contributions with uncertainty
Suppose interested in contributions by security
◦ let w = portfolio weights, C = covariance between securities
vk = variance contribution from security k
= cov(holdings in k, portfolio) = wk ∑j Ckj wj
Say parameters are uncertain and weights shift
◦ vk is random
◦ E[vk] = E[wk ∑j Ckj wj] = ∑j E[wk Ckj wj]
◦ C, w independent → = ∑j E[Ckj]E[wk wj] = ∑j E[Ckj] (E[wk] E[wj] + cov[wk, wj])
◦ Cov[vk,vj], after distributional assumptions, involves means and covariances of w and C
◦ Note: Mean and covariance of standard deviation contributions, sk = vk / σP , can’t be easily computed
Instead, there is a notional center E[vk] / sqrt(E[σP
2]) and approximate covariance from linearization
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
Need to be modeled
- 16. Modeling uncertainty
w: Comes from a model of price dynamics over the horizon of interest, e.g., overnight or until
the next rebalance
C: One way is an uncertain covariance model as in
◦ Shah, A. (2015). Uncertain covariance models. http://ssrn.com/abstract=2616109
◦ http://www.slideshare.net/AnishShah23/uncertain-covariance-62814740
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
- 17. Factor-modeled covariance in pictures
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
GOOG
Market
Risk Factors
Exposures
Stock-Specific
Effect
AAPL
GE
Growth-Value
Spread
Together these constitute a
model of how securities move –
jointly (winds and sails) and
independently (motors)
- 18. ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH
GOOG
Uncertain Risk Factors
projected onto orthogonal directions
Uncertain
Exposures
Uncertain
Stock-Specific
Effect
Factor variances have a mean and
covariance
Exposures have a mean and
covariance
Stock-specific effects have a
mean and variance
Calculated quantities – e.g., risk
contributions, portfolio variance
– have a mean and variance
Uncertain factor-modeled covariance
- 19. Summary
Risk decomposition is a canonical, useful tool to assess a portfolio’s risks
But perfect alignments (as arise from optimization) hide risks
… and ingredients are uncertain, so such alignments don’t really exist
… and price movement breaks them even if they did
The solution is to analyze with parameters and weights explicitly modeled as uncertain
Surfaces latent risks and fragility
Ideas are relevant to other risk applications
◦ For more stability, what about optimizing a portfolio’s expected future variance?
◦ or penalizing uncertainty during optimization
ANISHRS@INVESTMENTGRADEMODELING.COM
© 2021 ANISH SHAH