大浩浩的笔记课堂——FRM考试学习笔记117

Topic 46 Model Risk

1.The introduction of model risk:

⑴definition:

Model risk associated with using financial models to simulate complex relationship.

⑵importance:

①Model risk becomes important when quantifying the risk exposures of complex financial instruments,including exotic or synthetic derivatives and structured products.

②Model risk can give rise to losses from model errors,errors in assumptions,carelessness,fraud,or intentional mistakes.These errors can lead to undervaluing risk,overvaluing profit,or both.

⑶sources:

The sources of model risk is including:

①common model errors:

A.assuming constant volatility:

The most frequent error in model building is to assume that the distribution of the underlying asset stationary.

B.assuming a normal distributionof returns:

It will oversimplify a model.

C.underestimate the number of risk factors that must take into account

D.assuming perfect markets and adequate liquidity:

a/ Models are almost always derived under the assumption that perfect capital markets exist.

b/ liquidity or rather the absence of liquidity

E.misapplying a model:

a/ mathematically correct and useful,but misapplied to a given situation

b/ Models may not perform well when applied to subtly different instruments.

②common model implementation errors:

A.definition:

It is the risk associated with using financial models to simulate complex relationships.

B.occurrence:

They are occurred when models that require complex simulations are not allowed to run a sufficient number of runs(Monte Carlo simulation).This may result in incorrect output and therefore an incorrect interpretation of results.

C.problems:

The most frequent problems in estimating values:

a/ inaccurate data

b/ inappropriate length of sampling period:

"Old" data can become irrelevant and may introduce noise into the estimation process,as well as,putting higher weight on obsolete information.

c/ problems with liquidity and the bid/ask spread

D.considerations:

a/ frequency of refreshing model parameters(volatilities and correlations)

b/ correctly estimating parameters(durations,volatilities,correlations)

③common valuation and estimation errors:

A.inaccurate data

B.incorrect sampling period length:

increasing the length of sampling period→put higher weight on obsolete information

C.liquidity and valuation problems

④errors can be introduced into models:

A.programming error

B.implementation error

C.valuation risk:

Model errors in securities valuations or in hedging lead to market risk and operational risk.

2.Mitigating the model risk:

⑴To invest in research to improve models and to develop better statistical tools.

⑵To establish a process for the independent vetting of how models are both:

①documentation(assumption+mathematical formulas)

②soundness of model(reasonable representation:accepting/rejecting model)

③independent access to financial rates(independent parameter estimation)

④benchmark modeling(e.g.testing models against simulation)

⑤health check and stress test the model:

A.health check:

The model contains all of the necessary properties.

B.stress test:

Using simulations to check the model´s reaction to different situations.

⑥To incorporate model risk into the risk management risk:

Building a formal treatment of model risk into the overall risk management procedures,and periodically reevaluate models.

⑶To periodically revalue the models for relevance and accuracy:

Empirical evidence suggests that simple,robust models work better than more complex and less robust models.

⑷The fund relied on a VaR model:

①Using a 10-day horizon is too short to sufficiently model the time to raise new capital. 

②Do not factor in liquidity risk due to assuming perfect liquid.

③Do not incorporate correlation and volatility risks because the fact market has strong positive correlation.

大浩浩的笔记课堂之FRM考试学习笔记合集