Upload
yamanote
View
581
Download
0
Embed Size (px)
DESCRIPTION
On why a multi-factor or structural model of risk might be a good idea at the enterprise level, rather than the more common VaR models based simply on historical returns
Citation preview
D R A F T – W O R K I N P R O G R E S S
April 12th 2011Presented by Nick WadeDirector, Asia Marketing
Northfield Information Services [email protected]
+81(0)3 5403-4655+61(0)2 9238-4284
NORTHFIELDUsing a Structural Model for Enterprise Risk
D R A F T – W O R K I N P R O G R E S S
2
Talking Points for Today
Fund managers need detailed risk attribution for risk budgeting, performance attribution, scenario analysis and portfolio construction
CRO/Board/Trustees/Investment Committee need high-level risk across the firm
Client needs vary - Who is the asset owner?
Returns for illiquid assets e.g. direct property contain “appraisal smoothing” effects and do not reflect underlying risk i.e. it’s not a real price.
Risk Models typically used for fund management are not consistent across asset classes and markets, and usually ignore “difficult” asset classes like property, infrastructure, private equity.
We need to provide an integrated platform that can provide different levels of detail to different audiences, and include the “difficult” asset classes
D R A F T – W O R K I N P R O G R E S S
3
More about that split…
A split in the world:
– Portfolio Management and Asset Management Firms
Good: Risk analysis and portfolio management done using multi-factor models, contributions to risk aligned with contributions to return
E.g. what will happen if oil prices go up? Does my portfolio reflect my belief that Toyota will perform twice as well as BHP?
Bad: Typically single-asset class models; delivered in stand-alone software– Enterprise Risk
Good: the typical VaR measures are intuitive Good: typically an integrated platform Bad: one dimensional, return-based; problematic for Illiquid instruments
Solution: Find an integrated Platform that can do both
– Offer a customizable reporting tool on top of a set of smart components that can deliver both “VaR” and also multi-factor risk decomposition consistently and comparably across asset classes and markets satisfying both fund management and enterprise risk requirements
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Why VaR and return-based models may not be enough Many “enterprise” risk systems assume:
– Horizon: that the “enterprise” runs a risk of not surviving past the weekend e.g. 10 day VaR
– Returns-based: that the “enterprise” only invests in liquidly exchange traded instruments, and that therefore the observed returns tell you all you need to know about risk
no direct property, infrastructure, private equity, timber and so forth difficult assets can be “proxied” with equity securities
– Credit Ratings, Accounting data are reliable
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Why do we have this split?
Bank/Hedge Fund/Trading Desk– Risk management is about maintaining solvency in the event the value of
the assets declines i.e. Survival– Some measures of drawdown risk such as parametric VaR, are just the
standard deviation times a scaling factor. This is what the RiskMetrics system was created for at JP Morgan.
Asset Manager/Pension Fund/Sovereign Wealth– Risk is about how the variability of return from period to period reduces
compounding of returns over time. The geometric mean return of a portfolio is equal to the arithmetic mean of the returns minus one half the variance of the returns
– Investor wealth depends on the compounding of returns over time, so the variance of the returns (not the standard deviation) is the predominant issue in controlling investor risk. As such, the most relevant measure of risk to decompose is the variance, not the tracking error or VaR.
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
A Third Choice?
Create a multi-asset class risk model based on familiar approaches in asset management
Cover unlisted assets using a structural model, not historical returns
Use market-implied credit measures, not agency ratings
Embed into a firm-wide integrated platform
Offer reporting at different levels of granularity from enterprise VaR down to single-security contributions to (sub)portfolio risk, satisfying different constituencies
D R A F T – W O R K I N P R O G R E S S
77
Northfield Products and Services
Delivered as stand-alone applications or smart components within a partner integrated platform:
Portfolio Construction Tools (Optimizer, MARS)
Risk Management (Risk Model Family)
Asset Allocation Tools (ART)
Performance Attribution
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
What Northfield EE Can Doa multi-factor model of risk across asset classes
Analyze risk consistently across a broad range of asset classes Structural Model for unlisted/illiquid asset classes – avoid issues
with appraisal smoothing Structural Model for Credit Risk – avoid agency credit ratings Over 5 million individual instruments plus client defined terms:
– Global equity securities: developed and emerging, including externally managed funds
– Global fixed interest: developed and emerging government, corporate, convertible, CMO, MBS, ABS, muni, and agency bonds, and private placements, including externally managed funds
– Global REIT securities– Direct investment in property down to tenant level– Direct investment in infrastructure projects down to cashflow level– Derivative securities such as swaps, futures, options– Hedge Funds with undisclosed constituents
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
How Does Northfield EE do this?
NOT an unstable unclear roll-up of a dozen or more individual models
Atomic Approach to Risk One set of factors for all asset classes
– Market Risk As a function of macro- and micro-economic effects
– Interest rate risk The effect on the price of securities of yield curve shape changes
– Credit risk As a function of macro- and micro-economic effects
– Currency risk– Effect of embedded optionality– Effect of prepayments– Exposure of cashflows to macro- and micro-economic drivers of risk
Dynamic capture of new factors & transient effects via an adaptive hybrid risk model factor structure
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Combine macro, micro, and statistical factors
Gain the advantages of each, whilst mitigating the limitations of each
– Intuitive, explainable, justifiable observable factors– Minimal dependence on accounting information– Rapid inclusion of new or transient factors
Estimate using time-series approach– Diversify away estimation error– best for markets with moderate to low dispersion– best for portfolios with moderate/high diversification
Northfield Innovation – The Hybrid Model
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Equity in the EE framework
Macro factors– Market, sector, oil, interest rates, market development, currency
Micro factors– Size, value, growth, momentum
Temporary factors– Statistical factor analysis acting on the residual return to ensure full capture of
pervasive effects at all times
Adjustments for non-stationarity– Exponential weighting, conditional variances, Parkinson range measures, cross-
sectional dispersion adjustment
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business 12
Fixed Income in the EE framework
Interest rate risk– All bonds are priced relative to their local curves and their sensitivity to yield
curve shape changes is captured
Credit risk– market-derived measures independent of credit ratings– We expose the drivers of credit risk: macro, micro and issuer specific
Currency risk– Global currencies are explicitly included in the model
Optionality, Prepayment, Convertibles– Binomial tree pricing enables us to capture the effect of put, call, sinking fund,
and prepayment risk at each node on the tree– Quadrinomial (3D tree) allows accurate pricing of converts without Black-
Scholes assumptions
Issuer-specific risk– Individual bond risk also includes issuer-specific risk
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Additional Coverage:
Hedge Funds, undisclosed constituents: Return series and details provided by client or Northfield to our EENIAC tool Sharpe style-analysis to get index weightings Constituents optimized to reduce number of names to provide appropriate
specific risk
Derivative Securities Terms and conditions are provided by the client to our EENIAC tool A set of exposures and all other necessary files is returned to the client (for
example composite assets reflecting underlying basket)
Unlisted e.g. Direct Property Provide cashflow (lease-level) terms and conditions in prefabricated templates to
EENIAC Risk exposures and all necessary supporting files are returned (for example,
composite assets reflecting underlying holdings)
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Northfield EE Summary
EE provides: a consistent framework for enterprise risk and performance attribution 60,000 equity and 500,000 fixed interest instruments as standard structural model for credit risk that is independent of credit ratings structural model for illiquid assets avoiding appraisal smoothing On-demand access to a further universe of over 5 million CMO, MBS,
ABS, muni, and agency securities On-demand analysis of client-supplied instruments
Northfield EE is Available as a stand-alone application or via industry leading partners Integrated within DST Anova
INVESTMENT MANAGEMENT SOLUTIONS | Optimising Your Investment Business
Appropriate Models for Each Market
Investment sense, broad acceptance & academic approval
Intuitive factors used to describe risks
Analysis that is comparable across markets
Systematic risks detected & understood
“Glass box” – open, clear models & systems
Northfield risk analysis offers…