interest rates after the credit crunch...case study of ava mpu calculation for a security. • top...
TRANSCRIPT
![Page 1: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/1.jpg)
Prudent ValuationBridging the gap between pricing and risk management
Cass Business School
London, 4 November 2015
Marco Bianchetti – Intesa Sanpaolo, Financial and Market Risk Department
In collaboration with Umberto Cherubini – Associate Professor of Mathematical Finance, Univ. of Bologna
![Page 2: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/2.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 2
Summary
1. Introduction
2. Regulation
3. AVA calculation
4. Conclusions
5. References
![Page 3: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/3.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 3
1: IntroductionOverview
Traditionally, quantitative finance practitioners are divided into two populations: thosewho seek fair values, i.e. means of price distributions, and those who seek riskmeasures, i.e. quantiles of price distributions. Fair value people and risk people typicallylive in separate lands, and worship different gods: the profit and loss balance sheet, andregulatory capital, respectively.
Prudent Valuation is a rather unexplored midland which has recently emergedsomewhere in between the well known mainlands of Pricing and Risk Management. Infact, the Capital Requirements Regulation (CRR), requires financial institutions to applyprudent valuation to all fair value positions. The difference between the prudent valueand the fair value, called Additional Valuation Adjustment (AVA), is directly deductedfrom the Core Equity Tier 1 (CET1) capital. On March 31st 2014, the European BankingAuthority (EBA) published a draft Regulatory Technical Standards (RTS) for prudentvaluation, to be approved by the EU Commission.
The 90% confidence level required by regulators for prudent valuation links quantiles ofprice distributions (exit prices) to capital, thus bridging the gap between the Pricing andRisk Management mainlands, and forcing the crossbreeding of the fair value and riskpopulations above.
In this seminar, we will explore the Prudent Valuation land.
![Page 4: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/4.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 4
1: IntroductionOverview
Q-LandQ-measure
Pricing: extrapolate the presentFair value
Profit and loss
P-LandP-measure
Risk: model the future
Risk measuresCapital
Prudent LandPrudent measurePrice distribution90% exit price
Capital
See A. Meucci, “P versus Q: Differences and Commonalities between the Two Areas of Quantitative Finance”, GARP Risk Professional, pp. 47-50, February 2011, http://ssrn.com/abstract=1717163
![Page 5: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/5.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 5
1: Introduction Prudent valuation history [1]
August 2008
FSA “Dear
CEO letter”
November 2010
FSA “Product Control
Findings and Prudent
Valuation Presentation”
April 2012
FSA “Regulatory Prudent
Valuation Return”, Policy
Statement
2008 2009 2010 2011 20122006 20072004 2005
June 2004
BCBS “International Convergence
of Capital Measurement and
Capital Standards” (Basel 2),
sec. VI (“Trading book issues”),
ch. B (“Prudent valuation
guidance”), par. 690-701.
![Page 6: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/6.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 6
1: Introduction Prudent valuation history [2]
13 November 2012
EBA Discussion
Paper
(EBA/DP/2012/03)
10 July 2013
EBA Consultation
Paper
(EBA/CP/2013/28)
1 January
2014
CRR
575/2013
31 March 2014
EBA Final Draft RTS and first
application of prudent valuation
Q3 or Q4 2015
Expected final approval
of EBA RTS by the
European Commission
8 November 2013
EBA Quantitative
Impact Study
2012 2013 2014 2015
23 June 2015
EBA Final Draft
RTS amended
Dashed = unofficial
Prudent valuation in
place
![Page 7: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/7.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 7
2: Regulation CRR 575/2013 [1]
![Page 8: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/8.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 8
2: Regulation CRR 575/2013 [2]
Art. 34Prudent valuation
scope
Systems and
controls
Valuation
Valuation
adjustments
Art. 105
CRR
575/2013
CRR Prudent Valuation Tree
Prudent valuation
principles
Degree of certainty, art. 105.1
S&C requirements, art. 105.2
Revaluation frequency art. 105.3
Mark to market, art. 105.4-5
Mark to model, art. 105.6-7
IPV, art. 105.8
Valuation adjustments, art. 105.9-10
Illiquid positions, art. 105.11
Other valuation adj., art. 105.12
Complex products, art. 105.13
![Page 9: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/9.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 9
2: RegulationEBA RTS: overview
General
provisions
Sec. 1
Core
approach
Sec.3
EBA RTS
Final draft
EBA RTS Prudent Valuation Tree
Simplified
approach
Sec.2
Documentation
systems &
controls
Sec.4
Methodology for AVA, art. 1
Definitions, art. 2
Sources of market data, art. 3
Conditions of application, art. 4
AVA calculation, art. 5
AVA aggregation, art. 6
Overview, fall back, art. 7
General provisions, art. 8
AVA calculation, art. 9-17
Documentation, art. 18
Systems & controls, art. 19
Entry into force, art. 20 AVA OpR, art. 17
AVA EaT, art. 16
AVA FAC, art. 15
AVA CoPo, art. 14
AVA FVA, art. 13
AVA CVA, art. 12
AVA MoRi, art. 11
AVA CoCo, art. 10
AVA MPU, art. 9
![Page 10: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/10.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 10
2: RegulationEBA RTS: prudent valuation scope
Positions subject
to prudential
filters (AFS)
Positions in
hedge
accounting
Positions for which a
change in their
accounting fair value
has only a partial or
zero impact on CET 1
(art. 4.2 and 8.1)
EBA RTS Prudent Valuation scope: exclusions
Positions in
back to back
EU Gov. bonds
Other bonds
Equity
General criteria
for exclusionPositions excluded
% of
exclusion
100%
Partial
Partial
Simplified appr.
Partial, residual exposure
of hedged + hedging items
Core appr.
100%
Partial, residual exposure
to CVA, FVA, CoPo, FAC,
EaT, OpR AVAs
![Page 11: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/11.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 11
2: RegulationGlobal view of key regulatory concepts
Fair
value
CRR art. 34, 105
EBA RTS
Prudent
value
Scope and exclusions
90% confidence level
Simplified approach
Mark to market
Mark to model
IPV
Systems
and
controls
Core approach
Expert based
Fall back
Diversification
0.1% Formula
9 AVAs
![Page 12: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/12.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 12
Price distribution, fair value, fair value adjustment, prudent value, AVA
What about realprice distributions...?
Fair value
(mean)Fair value
adjusted
Prudent value
(quantile)
Fair value adjustment
AVA
2: RegulationFair Value Vs Prudent Value [1]
![Page 13: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/13.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 13
2: Regulation Fair Value Vs Prudent Value [2]
Fair Value
o Regulation: IFRS13
o Application: balance sheet
o Percentile: 50% (expected
value)
o The price that would be received
to sell an asset or paid to
transfer a liability in an orderly
transaction between market
participants at the measurement
date
o Must include all the factors that
a market participants would use,
acting in their economic best
interest.
o Atoms: single trades.
o Fair value adjustments
o Non-entity specific
Prudent value
o Regulation: CRR/EBA
o Application: CET1
o Percentile: 90%
o Must reflect the exit price at which
the institution can trade within the
capital calculation time horizon.
o Atoms: valuation positions subject
to a specific source of price
unertainty
o Entity specific
o Subject to diversification benefit
(50% weight for MPU, CoCo, MoRi
AVAs)
![Page 14: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/14.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 14
The EBA conducted a QIS to estimate the total impact of the requirements of the RTSincluding 59 banks across 15 jurisdictions, with the following results.
Small banks: < 15 €/bln Medium banks: 15 - 100 €/bln Large banks: > 100 €/bln
Average
227 €/mln
per bank
3: RegulationEBA RTS QIS Nov. 2013
![Page 15: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/15.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 15
Summary
1. Introduction
2. Regulation
3. AVA calculation
4. Conclusions
5. References
![Page 16: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/16.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 16
3: AVA calculationAVAs overview
Core approach
Additional Valuation Adjustments
Market
Price
Uncertainty
(MPU)
Art. 9
Close Out
Costs
(CoCo)
Art. 10
Model Risk
(MoRi)
Art. 11
Unearned
Credit
Spread
(CVA)
Art. 12
Investing &
Funding
Cost
(FVA)
Art. 13
Concen-
trated
Positions
(CoPo)
Art. 14
Future
Admin
Costs
(FAC)
Art. 15
Early
Termination
(EaT)
Art. 16
Main
AVAs
CVA/FVA
AVAs
Other
AVAs
Operational
Risk
(OpR)
Art. 17
The AVA hierarchy
Market risk factors
50% weights for diversification
Market risk factors
Split onto main AVAs
Non-market risk factors
100% weights, no diversification
![Page 17: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/17.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 17
3: AVA calculationDefinitions and basic assumptions
Definitions summary
Item Definition Comments
Fair value 𝐹𝑉 𝑡 =
𝑖=1
𝑁𝑝
𝐹𝑉𝑖 𝑡 i = index for valuation positions
Prudent Value𝑃𝑉𝑖𝑗𝑘 𝑡 ≤ 𝐹𝑉𝑖 𝑡
∀ 𝑖 = 1, … , 𝑁𝑝, 𝑗 = 1, … , 𝑁𝑢, ∀ 𝑘 = 1,… , 𝑁𝐴𝑉𝐴
o j = index for risk factors
o k = index for AVAs
Additional
Valuation
Adjustment
(simplified)
𝐴𝑉𝐴 𝑡 = 0.1%
𝑖=1
𝑁𝐴𝑠𝑠𝑒𝑡𝑠
𝐹𝑉𝑖 𝑡 +
𝑖=1
𝑁𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠
𝐹𝑉𝑖 𝑡𝐴𝑉𝐴 𝑡 is the total valuation
adjustment at time t
Additional
Valuation
Adjustment
(core)
𝐴𝑃𝑉𝐴𝑖𝑗𝑘 𝑡 ∶= 𝑤𝑘 𝐹𝑉𝑖 𝑡 − 𝑃𝑉𝑖𝑗𝑘 𝑡 ,
𝐴𝑉𝐴𝑘 𝑡 : =
𝑖=1
𝑁𝑝
𝑗=1
𝑁𝑢
𝐴𝑃𝑉𝐴𝑖𝑗𝑘 𝑡
o 𝐴𝑃𝑉𝐴𝑖𝑗𝑘 𝑡 is the k-th AVA
associated to source of
valuation uncertainty j and
valuation position i at time t,
o 𝐴𝑉𝐴𝑘 𝑡 is the total k-th AVA at t
Prudent
Valuation
Adjustment
𝑃𝑉𝐴 𝑡 ≔
𝐴𝑉𝐴(𝑡) Simplified
𝑘=1
𝑁𝐴𝑉𝐴
𝐴𝑉𝐴𝑘 𝑡 Core
𝑃𝑉𝐴 𝑡 is the total valuation
adjustment at time t
![Page 18: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/18.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 18
3: AVA calculationAVA Market Price Uncertainty (MPU)
Case study of AVA MPU calculation for a security.
• Top left: market bid and ask prices. FV is computed as average mid price = 162.25.
• Bottom left: ranking and percentiles of mid prices, AVA MPU for long and short positions, equal to 0.14 and 0.12, respectively.
• Top right: distribution chart.
![Page 19: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/19.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 19
AVA definitionAVA Model Risk (MoRi) refers to the valuation uncertainty of a valuation exposure arising from uncertainty in models and model calibrations used by market participants. In particular, AVA MoRi does not refers to the uncertainty in market risk capital arising from model risk (see FAQ 23.1).
AVA main referenceso EBA RTS, article 11. o EBA FAQs 10, 23.1, 28.
AVA scope of applicationWithin the general prudent valuation scope (see before), AVA MoRi refers in particular to those valuation positions for which the Institution estimates that there is a lack of firm exit price due to model and/or model calibration choices. Of course, instruments which can be replicated by exact static combination of mark-to-market instruments should not contribute to AVA MoRi.
3: AVA calculationAVA Model Risk (MoRi) [1]
![Page 20: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/20.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 20
Alternative models and calibrationsAVA MoRi is not based on any possible alternative model or model calibration, but on those specific alternative models or model calibrations that may reasonably used by market participants to price the same or similar valuation exposures.
Examples
o alternative but reasonable models, • calibrated to the same calibration basket
• Referred to the same group of financial instruments
o Same model, alternative calibration approaches, e.g. • different calibration baskets
• different calibration weights (e.g. flat, or vega weighted)
• different objective functions
• different optimization algorithm (e.g. global vs local)
• Etc.
o Same model, same calibration, alternative numerical approaches, e.g. • analitycal approximations
• semi-analitycal approximations
• numerical PDE solution
• Monte Carlo simulation
• etc.
3: AVA calculationAVA Model Risk (MoRi) [2]
Inspiration: «There’s
plenty of room at the
bottom»Richard Feynman, 1959
www.its.caltech.edu/~feynm
an/plenty.html
![Page 21: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/21.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 21
3: AVA calculationAVA MoRi: historical sources of model risk
Historical sources of model risk
Period Main driver Main risk factor Effects1987 Black Monday Volatility Volatility smile
2004 CMS market VolatilitySwaption volatility smile and CMS convexity adjustment
2004 IAS39 Credit Credit Risk Adjustment (CRA)2007 Credit crunch Credit, liquidity Subprime writedown2007 Credit crunch Interest rate basis Multiple yield curves
2009-2010 Credit crunch Interest rate basis CSA discounting2009-2010 Credit crunch Bilateral credit CVA & DVA (IFRS13, 2013)2013-2015 Credit crunch Funding Funding Valuation Adjustment (FVA)
2013-2014 Credit crunch Interest rateNegative interest rates and inflation, negative Floor strikes, Bond floater coupons floored, end of Black’s model.
2014- Credit crunch Capital charges Capital Valuation Adjustment (KVA)2017 Credit crunch Funding Bilateral initial margins
![Page 22: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/22.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 22
Market Risk Scenarios vs Model Risk Scenarioso Risk measures are typically linked to scenarioso Scenarios are related to the risk factors relevant for a particular risk typology
3: AVA calculationAVA MoRi: model risk scenarios vs traditional scenarios
Risk class Scenarios Risk measures
Market risk Present market data VaR, Expected shortfall, etc.
Counterparty risk Future market data EPE, Effective EPE, etc.
Operational risk Operational loss event frequency
and severity
VaR 99.9%
Model risk Model scenarios
o Alternative models
o Alternative numerical approaches
o Alternative calibrations
K-th percentile of distribution
of model prices (10°
percentile for Prudent
Valuation)
![Page 23: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/23.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 23
3: AVA calculationAVA MoRi: model risk scenarios for interest rate derivatives
![Page 24: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/24.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 24
3: AVA calculationAVA Model Risk (MoRi): case study 1 [1]
Case study 1: model risk in interest rate yield curve construction
Interest rate yield curves are used everywhere for discounting and for interest rate derivatives and securities with floating rate coupons. So, this is an important case study.
Yield curve construction is based on recursive application of pricing formulas applied to interest rate market instruments. So, there is a lot of modelling inside.
In particular, the interpolation algorithm is very important, both pre and post bootstrapping:
o Simple but non-smooth linear interpolation algorithms are very simple and robust, but produces irregular forward curves
o Standard spline interpolation is less simple but produces oscillating yield curves
o Monotonic cubic spline interpolation is regular.
![Page 25: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/25.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 25
3: AVA calculationAVA Model Risk (MoRi): case study 1 [2]
Linear interpolation on zero interest rates
Monotonic cubic spline interpolation on zero interest rates
![Page 26: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/26.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 26
3: AVA calculationAVA Model Risk (MoRi): case study 1 [3]
Differences in bps between three different interpolation algorithms (linear, natural cubic spline and monotonic cubic spline) for a portfolio of 3 standard IRS on Euribor 1M, 6M, 12M + 3 standard Basis Swaps.
![Page 27: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/27.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 27
3: AVA calculationAVA Model Risk (MoRi): case study 2 [1]
Case study 2: model risk experiment with Numerix
Sensitivity of prices to modelso Various dimensions of modelling decisionso Example of Bermudan swaption pricing with HW1F, HW2F, CIR, and BK modelso Impact of calibration choiceso AVA MoRi for a Bermudan swaptiono Model implied European swaption smile
Impact of changing market environment on model performanceo Handling of negative rateso Example of floor pricing with very low strikes by using various models
Joint work with Ilja Faerman and Laure Darleguy, Numerix webinar, 12 Nov. 2014, available at www.numerix.com
![Page 28: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/28.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 28
3: AVA calculationAVA Model Risk (MoRi): case study 2 [2]
Case study 2: model risk experiment with Numerix (cont’d) Global modelling approach
Trade
FX spotBasis
spreadYield Curve Correlation
Model
underlying
Forward
curveSwap rate
Risk factor
Short-rate
Distribution
typeNormalLog-normal Mixture
Chi-
squared
Model type HW1F HW2F
Calibration
instruments Caplets Swaptions
Instruments
configuration
10Y
diagonal
20Y
diagonal10Y column
10Y diag +
10Y column
CIR BK
CMS
…
…
…
…
…
…
![Page 29: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/29.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 29
3: AVA calculationAVA Model Risk (MoRi): case study 2 [3]
Case study 2: model risk experiment with Numerix (cont’d)
Experiment
#Instruments Models Calibrations
Bermudan
swaption
• Coterminal bermudan payer
swaption
• Euribor 6M
• 10Y maturity
• Annual callability
• Sstrike ATM 10Y swap
• OIS discounting
• Hull-White 1 Factor
(HW1F)
• Black-Karasinski (BK)
• Cox-Ingersoll-Ross 1
Factor (CIR1F)
• Hull-White 2 Factors
(HW2F)
• Cox-Ingersoll-Ross 2
Factors (CIR2F)
• Set 1: 10 Y diagonal
swaption ATM
• Set 2: 10Y diagonal
and 1Y column
swaption ATM
• Set 3: 20Y diagonal
and 1Y column
swaption ATM
Caps/Floors
with negative
rates
• 5Y Floor
• Euribor 6M
• Negative and positive strikes
• Yield curves with negative
rates
• Linear interpolation and flat
extrapolation
• SABR interpolation and flat
extrapolation
• Black (analytic)
• Hull-White 1 Factor
(HW1F)
• Shifted Black-Karasinski
(SBK)
• Set 1: Cap volatility
columns for strikes
ATM and 1%
• Set 2: full Cap volatility
surface, with strikes
from 1% to 10%
Joint work with Ilja Faerman and Laure Darleguy, Numerix webinar, 12 Nov. 2014, www.numerix.com
![Page 30: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/30.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 30
3: AVA calculationAVA Model Risk (MoRi): case study 2 [4]
Overview of results
Prices range from 1.45% to 3.91% Normal models produce consistently higher PVs for all calibration sets compared to
non-normal models
HW1FBK
CIR1FHW2F
CIR2F
0.00%
1.00%
2.00%
3.00%
4.00%
Set1
Set2
Set3
Bermudan swaption prices per model and calibration set
![Page 31: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/31.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 31
3: AVA calculationAVA Model Risk (MoRi): case study 2 [5]
Results by calibration set
Calibration set 1 (10Y diagonal) produces highest distribution of prices Average price is fairly stable across different calibration sets Same model stays consistently below or above the average price for all calibration
sets
1.00%
1.50%
2.00%
2.50%
3.00%
3.50%
4.00%
4.50%
Set1 Set2 Set3
Bermudan swaption prices per calibration set
HW1F BK CIR1F HW2F CIR2F Average
![Page 32: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/32.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 32
3: AVA calculationAVA Model Risk (MoRi): case study 2 [6]
Results by model
HW1F and BK models exhibit lowest variations in prices with changing calibration set Prices of 1F and 2F models of the same model type can differ significantly
1.00%
1.50%
2.00%
2.50%
3.00%
3.50%
4.00%
4.50%
HW1F BK CIR1F HW2F CIR2F
Bermudan swaption prices per model
Set1 Set2 Set3
![Page 33: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/33.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 33
3: AVA calculationAVA Model Risk (MoRi): case study 2 [7]
Results
Notional is 10m EUR Assuming Fair Value is the average of all price
Long swaption:o Fair Value: FV = 258k EURo Prudent value is the 10% percentile of all prices: PV = 177k EURo AVA MoRi = 0.5x(FV-PV) = 40.5k EUR
Short swaption:o Fair Value: FV = -258k EURo Prudent value is the 90% percentile of all prices: PV = -317k EURo AVA MoRi = 0.5x(FV-PV) = 29.5k EUR
![Page 34: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/34.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 34
3: AVA calculationAVA Model Risk (MoRi): case study 2 [8]
Excluding models
All models All except HW2F All models All except HW2F
Fair Value (1) 258 258 -258 -258
Prudent Value 177 158 -317 -315
Model Risk AVA 40.5 50 29.5 28.5
Long swaption Short swaption
Fair Value (1) is computed as the average of all model prices
Fair Value (2) for “All except HW2F” is computed excluding the price of the HW2F model
All models All except HW2F All models All except HW2F
Fair Value (2) 258 240 -258 -240
Prudent Value 177 158 -317 -315
Model Risk AVA 40.5 41 29.5 37.5
Short swaptionLong swaption
![Page 35: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/35.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 35
3: AVA calculationAVA Model Risk (MoRi): case study 2 [9]
Exercise probabilities
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
Oct-15 Oct-16 Oct-17 Oct-18 Oct-19 Oct-20 Oct-21 Oct-22 Oct-23
Call probabilities per couponCalibration set 1
HW1F
BK
CIR
HW2F
CIR2F
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
Oct-15 Oct-16 Oct-17 Oct-18 Oct-19 Oct-20 Oct-21 Oct-22 Oct-23
Call probabilities per couponCalibration set 2
HW1F
BK
CIR
HW2F
CIR2F
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
Oct-15 Oct-16 Oct-17 Oct-18 Oct-19 Oct-20 Oct-21 Oct-22 Oct-23
Call probabilities per couponCalibration set 3
HW1F
BK
CIR
HW2F
CIR2F
Exercise probability per coupon CIR-type models imply a higher
probability of early exercise than HW models
The term structure of exercise probabilities is regular for all models for calibration set 1, humped for calibration sets 2 and 3.
![Page 36: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/36.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 36
3: AVA calculationAVA Investing and Funding Costs [1]
AVA definitionAVA Investing and Funding Costs (FVA) refers to the valuation uncertainty in the funding costs used when assessing the exit price of a valuation position, according to the applicable accounting framework. Such valuation uncertainty refers, in particular, to MPU, CoCo and MoRiuncertainties in the calculation of the funding cost. Hence, AVA FVA shall be split into such components, to be aggregated to their corresponding AVAs.
AVA scope of applicationWithin the general prudent valuation scope (see before), AVA FVA refers in particular to those valuation positions subject to a funding valuation adjustment and specifically, to OTC derivatives. Securities are excluded, since funding risk is already included in the security credit spread
![Page 37: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/37.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 37
3: AVA calculationAVA Investing and Funding Costs [2]
FVA losses as of end of 2014.
See “The Black Art of FVA, part III: a $4 billion
mistake?“, Risk, 2 Apr. 2015.
![Page 38: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/38.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 38
3: AVA calculationAVA Investing and Funding Costs [3]
Joint work with Alexander Antonov, Ion Mihai, to appear in Risk, November 2015.
![Page 39: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/39.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 39
4: Conclusions
![Page 40: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/40.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 40
5: ReferencesRegulations [1]
1) BCBS, “International Convergence of Capital Measurement and Capital Standards – A
revised framework”, June 2004.
2) Financial Services Authority, “Dear CEO Letter: Valuation and Product Control”, August
2008, http://www.fsa.gov.uk/pubs/ceo/valuation.pdf
3) Financial Services Authority, “Product Control Findings and Prudent Valuation
Presentation”, November 2010, http://www.fsa.gov.uk/pubs/other/pcfindings.pdf
4) Financial Services Authority, “Regulatory Prudent Valuation Return”, Policy Statement
12/7, April 2012, http://www.fsa.gov.uk/library/policy/policy/2012/12-07.shtml
5) International Accounting Standards Board, «International Financial Reporting Standards
13 – Fair Value Measurment», 1° Jan. 2013, www.ifrs.org
6) Regulation EU N.575/2013 of the European Parliament and of the Council on prudential
requirements for credit institutions and investment firms and amending Regulation EU
N.648/2012, 26 June 2013
7) European Banking Authority, “Discussion Paper relating to Draft Regulatory Technical
Standards on prudent valuation under Article 100 of the draft Capital Requirement
Regulation (CRR)” EBA/DP/2012/03, 13 November 2012, http://www.eba.europa.eu/-
/eba-discussion-paper-on-draft-regulatory-standards-on-prudent-valuation.
![Page 41: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/41.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 41
5: ReferencesRegulations [2]
8) European Banking Authority, “Consultation Paper Draft Regulatory Technical Standards
on prudent valuation under Article 105(34) of Regulation (EU) 575/2013 (Capital
Requirements Regulation – CRR)”, EBA/CP/2013/28, 10 July 2013,
http://www.eba.europa.eu/regulation-and-policy/market-risk/draft-regulatory-technical-
standards-on-prudent-valuation.
9) European Banking Authority, “Questions and Answers on prudent valuation”, October
2013, http://www.eba.europa.eu/-/revised-faqs-on-prudent-valuation-q-1.
10)European Banking Authority, “Quantitative Impact Study on prudent valuation”,
November 2013, http://www.eba.europa.eu/-/eba-launches-qis-exercise-on-prudent-
valuation.
11)Bank of Italy, Circolare 285, “Disposizioni di vigilanza per le banche”, 17 December
2013, https://www.bancaditalia.it/compiti/vigilanza/normativa/archivio-
norme/circolari/c285/index.html
12)European Banking Authority, “EBA final draft Regulatory Technical Standards
Regulatory Technical Standards on prudent valuation under Article 105(14) of
Regulation (EU) 575/2013 (Capital Requirements Regulation – CRR)”, 31 March 2014,
https://www.eba.europa.eu/regulation-and-policy/market-risk/draft-regulatory-technical-
standards-on-prudent-valuation
![Page 42: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/42.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 42
5: ReferencesRegulations [3]
13)European Banking Authority, “EBA final draft Regulatory Technical Standards
Regulatory Technical Standards on prudent valuation under Article 105(14) of
Regulation (EU) 575/2013 (Capital Requirements Regulation – CRR)”, rev1, 23 January
2015, https://www.eba.europa.eu/regulation-and-policy/market-risk/draft-regulatory-
technical-standards-on-prudent-valuation
![Page 43: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/43.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 43
5: ReferencesPapers
1) Richard Roll, “A simple implicit measure of the effective bid-ask spread in an efficient
market”, The Journal of Finance, Vol. XXXIX, n. 4, Sept. 1984.
2) E. Derman, "Model Risk", Goldman Sachs Quantitative Strategies Research Notes, Apr.
1996.
3) R. Rebonato, "Theory and Practice of Model Risk Management”, Quantitative Research
Centre (QUARC) of the Royal Bank of Scotland, 2002.
4) R. Cont, "Model uncertainty and its impact on the pricing of derivative instruments",
Mathematical Finance, Vol. 16, No. 3, July 2006, 519–547.
5) R. Brar, “A Regulatory Perspective on Prudent Valuation and Best Practice in Product
Control”, in “Managing Illiquid Assets”, E. Takagawa editor, Risk Books, 2012.
6) Tanguy Dehapiot, “Prudent Value”, Risk Minds presentation, Dec. 2014.
![Page 44: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/44.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 44
5: ReferencesOthers
1) Ernst & Young, “Prudent Valuation”, 24 May 2013.
2) Ernst & Young, “BIS III – Prudent Valuation – AVAs Overview and relations to IFRS13”,
July 2013.
3) Deloitte, “Prudent Valuation”, August 2013, http://www.deloitte.com/assets/Dcom-
Belgium/Local%20Assets/Documents/EN/Insights/FSI/be-fsi-
prudentvaluation_ebaconsultationpaper_aug2013.pdf.
4) Financial Machineries, http://www.financial-machineries.com.
![Page 45: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/45.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 45
10 Dec. 2014: Risk Minds Conference, joint talk on prudent valuation with T. Dehapiot.
28 May 2014: London Stock Exchange, Milano, prudent valuation course, M.
Bianchetti, U. Cherubini, E&Y.
16 May 2014: ABI conference, Roma, talk “Funding Valuation and Prudent Valuation
Adjustments (PVA & FVA)”, M. Bianchetti, U. Cherubini
24 Sept. 2014: corso ABI, Milano, talk “Prudent valuation“, M. Bianchetti, P. Virgili.
12 Nov. 2014: webinar Numerix, “Prudent Valuation: Bridging the Gap Between
Pricing & Risk Management”, M. Bianchetti (link).
24 Nov. 2014: London Stock Exchange, Milano, prudent valuation course, M.
Bianchetti, U. Cherubini, E&Y.
10 Dec. 2014: Risk Minds, Amsterdam, talk “Prudent Valuation - Bridging Pricing And
Risk Management”, M. Bianchetti (link).
25 Mar. 2015: WBS 4th CVA conference, London, corso “Prudent valuation“, M.
Bianchetti, U. Cherubini (link)
May 2015: Global Derivatives, Amsterdam, talk “Prudent Valuation - Bridging Pricing
And Risk Management”, M. Bianchetti (link).
5: ReferencesEvents
![Page 46: Interest Rates After the Credit Crunch...Case study of AVA MPU calculation for a security. • Top left: market bid and ask prices. FV is computed as average mid price = 162.25. •](https://reader036.vdocuments.net/reader036/viewer/2022071402/60eec0eb70379765e9356607/html5/thumbnails/46.jpg)
M. Bianchetti - Prudent Valuation – Cass Business School– London, 4 Nov. 2015 p. 46
Disclaimer and acknowledgments
Disclaimer
The views and the opinions expressed here are those of the author and do not
represent the opinions of his employer. They are not responsible for any use that may
be made of these contents. No part of this presentation is intended to influence
investment decisions or promote any product or service.
Acknowledgments
The authors gratefully acknowledges
o E. Maffi, A. Pignataro, S. Vasconi, F. Bertolini, M. Benvenuti, S. Vella from E&Y for
their contribution to develop the prudent valuation framework.
o I. Faerman from Numerix for his contribution for model risk examples.
o T. Dehapiot for sharing information and experties on the subject.
o Members of the AIFIRM committee on market risk for the stimulating discussions on
prudent valuation methodology and applications.
o Many other colleagues in Front Office and Risk Management of Intesa Sanpaolo for
creating a fertile environment to grow the seeds of prudent valuation.