DASH SSRN Finance/Risk Papers 2018

Jan Dash
https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=566347
File = DASH_SSRN_FinanceRisk_papers_2018.docx
Jan Dash
Bloomberg LP
Formerly head risk quant
731 Lexington Ave
New York, NY 10022
United States
SCHOLARLY PAPERS
27
DOWNLOADS
Rank 21,984
1,934
Ideas:
“ Contact: jdash9@comcast.net ; J. Dash Consultants
LLC Editor, Encyclopedia of Climate Change and Finance,
Economics, Policy (World Scientific, to be published)
Book: Quantitative Finance and Risk Management - 2nd
Ed. with chapter on Climate Change Risk Management:
http://www.worldscientific.com/worldscibooks/10.1142/900
3 ”
Scholarly Papers (27)
1.
Advanced Idiosyncratic Risk and Multi-Factor Models

Number of pages: 24 Posted: 07 Apr 2015 Last Revised: 02 Feb 2017
Jan Dash and Mario Bondioli
Bloomberg LP and Bloomberg L.P.
Downloads200(138,250)
• View PDF
• Download
Abstract:
A common scenario risk analysis employs a multiple factor model with assumed changes
in the factors to obtain changes in non-factor variables. This analysis is sometimes
designated as a “predictive stress scenario”.
We choose to designate the factor model as a multifactor “CAPM” model, also related to
APT models. We are not concerned with the details of the factor model. The risks of the
scenarios are evaluated for a portfolio of instruments depending on changes of the
variables.
A systemic problem with this factor model approach is that the empirical correlations
between pairs of non-factor variables are not maintained. We designate “advanced
idiosyncratic risk” or “AI-Risk” as the correction to the risk of a predictive stress CAPM
model scenario that includes a better approximation to the physical correlations. We have
developed a new, effective, and economical AI-Risk formalism. There are two parts to
AI-Risk. The first part, which is actually standard, has independent normal random
variables with no correlations; we designate that as part of the CAPM model to get the
variances correctly reproduced; we believe this is standard practice. The second part
contains the corrections to the correlations; this is new.
We evaluate AI-Risk for some stock portfolios. We find that the AI-Risk can be
important, typically between 10%-50% of the predictive stress risk. The amount depends
on the ratio of long/short (“L/S”) position values, with larger AI-Risk for higher L/S
ratios.
A technical remark is that we use correlations as the metric for model improvement.
Therefore we work with z-scores or “unit” returns with the empirical volatility divided
out.
Finally, we indicate the generalization to cross-sectional regression factor models.
In summary this work achieves two results:

1. A parsimonious formalism incorporating correlated idiosyncratic risk that gives better
agreement with given empirical correlations (so risk is calculated closer to the physical
world).
2. Indications of substantial addition to risk due to correlated idiosyncratic contributions,
with respect to the smaller risk obtained in the standard model with uncorrelated
idiosyncratic risk.
2.
Psychology, Stock/FX Trading, and Option Prices

Journal of Behavioral Finance, Forthcoming
Number of pages: 36 Posted: 27 Jan 2012 Last Revised: 21 Apr 2014
Alan Beilis, Jan Dash and Jacqueline Volkman Wise
affiliation not provided to SSRN, Bloomberg LP and Temple University - Risk Management &
Insurance & Actuarial Science
• View PDF
• Download
•
Abstract:
The financial crisis of 2008 had many putative causes. Psychology was an important
driver for human decisions underlying these causes. However, quantitative financial
models have no “knobs” to dial psychology parameters, and so arguably cannot possibly
cope with financial crises. We have no illusions of the difficulty of including psychology
in financial modeling. Here we take a first step by considering how a particular aspect of
psychology can influence an underlying security and subsequent option prices, in a
quantitative model. The underlying security can be a stock or an FX rate. There are three
steps. First we investigate how psychological regret and fear impact trading selling
behavior. Second we use results from the first step to link this changed trading behavior
with induced changes in underlying security prices. Third, we consider changes in option
prices due to these induced underlying security price changes. The results can be
expressed either as a modified effective dividend for stock options, a modified effective
interest rate for FX options, or an unusual change in implied volatility. Options analysis
for some USDCAD FX European options with implied parameters indicates this
approach has some empirical relevance.
The contribution of this paper is thus threefold: 1. The paper breaks ground by
emphasizing the desirability of incorporating interdisciplinary explicit interaction
between behavioral finance and securities modeling, 2. The paper provides a definite
model with a quantitative mechanism of how a particular psychological behavior can
influence the prices of some securities, and 3. The paper shows that this model can
facilitate the description of some illustrative option data.
3.
Smart Monte Carlo, Path Integrals, and American Options

Number of pages: 19 Posted: 13 Jul 2016 Last Revised: 17 Sep 2016
Jan Dash and Xipei Yang
• View PDF
• Download
•
Abstract:
In two previous papers we introduced Smart Monte Carlo SMC, more accurate and faster
than traditional MC. Here we apply SMC to American Monte Carlo AMC. The main tool
is the Feynman-Wiener path integral with a useful binning procedure. We also suggest
Prony interpolating functions with regression advantages. We present American put
option results. We obtain a smaller price error and a smoother optimal exercise boundary
than with standard methods. We comment on multidimensional SMC.
4.
Market Crises, Earthquakes, and the Reggeon Field Theory

Number of pages: 22 Posted: 07 Jun 2013
• View PDF
• Download
•
Abstract:
This paper contains new results for helping to understand financial crises. First, we
present a new model for obtaining the probability of equity crises within one year in
advance, and we test it. Second and separately, various markets already in crises appear
quantitatively related to a theory of nonlinear diffusion called the Reggeon Field Theory
calculated years ago. Details are in a longer companion paper i.
The time dynamics of the origins of crises are commonly pictured by bubbles growing
and collapsing. Our dynamic model, the “CEEC (Critical Exponent Earthquake Crisis)
Model”, has these features , and can provide early warning equity crisis signals to help
prevent losses. The CEEC Model uses concepts of “critical exponents” from physics plus
a qualitative analogy from earthquakes to describe the build-up of bubbles (increase of
“frictional stress”) with subsequent crises from bubble collapses (“earthquakes”) . The
only inputs to the CEEC model for an equity index are equity index returns.
Tests comparing the CEEC model to data yield encouraging results, much better than
chance. Here are results from running the model as the user would have run it in the past:
• In 69% of the tests for which the model indicated a crisis in the short term (within one
year), a crisis was in fact observed within one year.
• 31% of actual crises were missed by the model within one year before the crisis.
We give a summary of the CEEC model in this short paper. A longer companion paper
describing full details of the model will appear.
We also analyze various markets already in crisis (equity, FX, commodities, rates,
bonds). We find behavior numerically consistent with a theoretical result with no free
parameters for the general theory of nonlinear diffusion in physics, generalizing standard
Brownian motion, the Reggeon Field Theory RFT. An anomalous RFT critical exponent
translated into finance language is around 0.3, and this number qualitatively describes the
average behavior of markets in crisis.
This 0.3 RFT anomalous variance exponent is perhaps the first number calculated in
advance since the Gaussian Brownian diffusion variance exponent (1.0) used in standard
finance, without numerically fitting anything.
5.
Cleaning Financial Data Using SSA and MSSA

Jan Dash and Yan Zhang
Bloomberg LP and Bloomberg LP
• View PDF
• Download
•
Abstract:
We introduce a powerful method for cleaning time series - Multi-Channel Singular
Spectrum Analysis (MSSA). “Cleaning” means filling data gaps and removing
unphysical spikes, which are chronic problems. MSSA utilizes all available information
in “time” and “space” with autocorrelations, correlations, and lagged correlations. MSSA
performs demonstrably better than other methods. Here we present tests using MSSA to
fill data gaps, with positive results. Spike removal is in a separate paper.
6.
Path Integrals and Greeks

Number of pages: 10 Posted: 17 Jul 2016 Last Revised: 02 Feb 2017
Jan Dash
Bloomberg LP
• View PDF
• Download
•
Abstract:
Path integrals are useful and general. Here we show how to calculate Greeks using path
integrals. In particular we exhibit the solution to an otherwise troublesome calculation –
gamma for a digital option. A subsequent paper will present more detail.
7.
Data Spike Cleaning with MSSA

Number of pages: 16 Posted: 20 Jul 2016 Last Revised: 11 Nov 2016
• View PDF
• Download
•
Abstract:
This paper introduces a powerful method for detecting and fixing unphysical spikes in
time series. The method utilizes Multiple Singular Spectrum Analysis (MSSA) to define
local market trends used to identify outlier data spikes that are not caused by market
movements, and then effectively correct the spikes. Various refinements for spike
identification are proposed. This work extends previous work using MSSA to fill gaps or
holes in time series.
8.
The Macro-Micro Model, Trends vs. Noise, and SSA - I

Number of pages: 12 Posted: 13 Jul 2016 Last Revised: 22 Jul 2016
Jan Dash, Xipei Yang and Mario Bondioli
Bloomberg LP, Bloomberg L.P. and Bloomberg L.P.
• View PDF
• Download
•
Abstract:
The Macro Micro (MM) model contains different time scales and deals with risk as it
occurs in the real world, especially trend risk. A new methodology - Singular Spectrum
Analysis (SSA) – is introduced to identify historical trends, trend volatilities, and noise-
reduced trend-trend correlations. Application is for a long-term PFE risk simulation,
including realistic quasi-random Macro trends. Two Macro versions of the MM model
are presented – a statistical trend model and a historic “time-slice” trend model.
9.
SSA, Random Matrix Theory, and Noise-Reduced Correlations

Jan Dash, Xipei Yang, Mario Bondioli and Harvey J. Stein
Bloomberg LP, Bloomberg L.P., Bloomberg L.P. and Bloomberg L.P.
• View PDF
• Download
•
Abstract:
This is the third paper in a series devoted to obtaining noise-reduced, stable correlations
by smoothing time series using Singular Spectrum Analysis, or SSA. Here we show that
the SSA-based correlations are superior in terms of noise reduction, employing a number
of simple tests using Random Matrix Theory (RMT) constructs. In each case, the
correlations obtained using SSA-smoothed time series are further from noise than are
conventional correlations. “Noise” here is defined by a zero-correlation Wishart random
matrix WRM composed of correlations between series filled with independent Gaussian
random numbers.
10.
Predicting Equity Crises, Critical Exponents, and Earthquakes -

II
• View PDF
• Download
•
Abstract:
We present further encouraging evidence for the Critical Exponent Earthquake Crisis
(CEEC) Model that gives the probabilities of equity crises one year in advance. The
CEEC model uses suitable precursor signals and is agnostic regarding dynamical origins
of crises. The precursors accumulate in time between crises, like precursors to some
earthquakes. The model uses a sophisticated noise filter to separate out crisis signals. The
main metric is the anomalous exponent of an equity series describing the difference of the
data variance scaling exponent from the Brownian variance scaling exponent of 1. No
extra non-equity variables are used. The CEEC Model results for predicting crises are not
perfect, but are much better than chance, including out-of-sample tests. The details here
supplement our previous CEEC crisis paper (2013).
In another paper (2016) we give details showing that various markets - not just equities -
that are already in crisis are on the average described by a critical exponent of the
nonlinear-diffusion Reggeon Field Theory (RFT), calculated in 1974, with no free
parameters. Nonlinear diffusion naturally extends Brownian motion. Rich/cheap crisis
behavior is suggested as a paradigm. This supplements our previous CEEC crisis paper
(2013).
The CEEC Model crisis predictions use the same scaling form described by the
anomalous exponent as does the RFT. This consistency for crisis predictions and crisis
behavior is significant.
11.
Cleaning Data with Real-World Updating Using MSSA

Number of pages: 14 Posted: 20 Jul 2016
• View PDF
• Download
•
Abstract:
Data cleaning in the real world has to cope with new data arriving (or failing to arrive) as
time passes, and which may be bad data. We illustrate MSSA data cleaning with a real-
time historical simulation on some problematic data. The example also serves to
determine some MSSA algorithm parameters. This paper expands previous work on
MSSA applied to data cleaning (filling data holes, removing spikes, simulation)
described in previous papers.
12.
Path Integrals and Smart Monte Carlo - I

• View PDF
• Download
•
Abstract:
We present “Smart Monte Carlo” or SMC, improving the efficiency of Monte Carlo
(MC) simulations. SMC has two “stages”. The first stage, run adaptively for each deal,
produces equivalent results to standard MC simulation using fewer calls to the time-
consuming pricing functions. The second SMC stage is standard MC simulation for a
portfolio, using price look-ups from the first SMC stage. The result has improved
accuracy and speed. Examples are given.
13.
Noise-Reduced Correlations, the Signal to Noise Ratio, and SSA

• View PDF
• Download
•
Abstract:
This is the second paper presenting noise-reduced, stable correlations for long-term risk
measurement. We smooth time series using Singular Spectrum Analysis (SSA) and then
form the correlations from these smoothed time series. These correlations have superior
time stability and are cleaned of noise. Here we show that the Signal-to-Noise Ratio is
larger for the SSA-based correlations than usual, and we perform other signal/noise tests.
We use new results refining random matrix correlations.
14.
Analytic Solution to the Two Dimension Merton Model

Jan Dash, Mario Bondioli and Harvey J. Stein
• View PDF
• Download
•
Abstract:
We present an exact analytic solution to the two-dimensional correlated default structural
Merton model in the form of a local volatility problem using a conformal square-root
transformation of the exact solution to a 2D hybrid barrier problem. We also give an
approximation and evaluate it numerically to give an example. Finally we give an exact
simpler solution for the zero correlation case.
15.
Macro-Micro, Trends vs. Noise, and SSA - II

Jan Dash, Xipei Yang and Mario Bondioli
• View PDF
• Download
•
Abstract:
We describe some details of extensions of the Macro Micro (MM) model. Applications
include a long-term real-world PFE risk simulation, including realistic quasi-random
Macro trends. The details elaborated here include the use of the 3rd order skew Green
function to obtain micro mean reversion, a random time distribution for the Macro
component, sliding down the yield curve, and approximate no-arbitrage in the MM
model.
16.
Path Integrals and Smart Monte Carlo - II

• View PDF
• Download
•
Abstract:
“Smart Monte Carlo” (SMC) improves accuracy and speed. We extend results in an
earlier paper, applying SMC to path-dependent deals and multifactor models. Auxiliary
results are a path-to-path distance, an analytic approximation for N-dimensional Gaussian
integrals, and time interpolation results. We suggest “model perturbation” using simple
approximate models, and introduce a new product “DIAS” for better mortgage servicing
hedging. We prove a consistency condition for the MRG interest-rate model, and discuss
real-world vs. risk-neutral simulations.
17.
A Distressed Bond Model

Jan Dash, Xipei Yang and Stan Maydan
Bloomberg LP, Bloomberg L.P. and Bloomberg LP
• View PDF
• Download
•
Abstract:
We present the framework for a distressed bond model. The utility is as a proxy for
calculating the risk of a distressed bond portfolio. We elaborate several possible
implementations and give an example.
18.
Introduction to Noise-Reduced Correlations Using Singular

Spectrum Analysis
Number of pages: 12 Posted: 30 Aug 2017
Jan Dash, Xipei Yang, Mario Bondioli and Harvey J. Stein
• View PDF
• Download
•
Abstract:
We summarize new results for estimating correlations for use in risk management. These
estimates have better behavior than traditional estimation approaches from both a
business standpoint and a technical standpoint. We smooth time series using Singular
Spectrum Analysis (SSA) and compute correlations based on these smoothed series. We
demonstrate that SSA-based correlation estimates have less noise than standard
correlation estimates between unsmoothed series using: the signal-to-noise ratio, and
distances from noise using polynomials generalizing the z-score and random matrix
theory constructs. New useful analytic estimates for all eigenvalues of a random matrix
are described. SSA-based correlations also enjoy superior time stability. Technical
aspects are given in four accompanying papers, including extensive analyses of time
stability and the noise-reduction tests described in this short paper.
19.
Speeding Up VAR with Smart Monte Carlo

Number of pages: 34 Posted: 01 Mar 2018
Jan Dash and Xinchong Zhang
• View PDF
• Download
•
Abstract:
We show that VAR calculation speedup of an order of magnitude can be obtained using
Smart Monte Carlo with a sophisticated interpolator. As a byproduct, we give some
encouraging numerical results for evaluating N-dimensional Gaussian integrals without
doing any integrals at all.
20.
Advanced Idiosyncratic Risk and Multi-Factor Models – Short

Version
Number of pages: 19 Posted: 01 Feb 2017
• View PDF
• Download
•
Abstract:
We introduce advanced idiosyncratic risk (“AI-Risk”), a parsimonious correlated residual
correction to a predictive stress CAPM-like factor model, aimed to get more accurate
stock-stock correlations. We find that AI-Risk can be significant for stock portfolios.
Inclusion of AI-Risk gives a more realistic risk assessment, consistent with real-world
correlation constraints. We also indicate the generalization of AI-Risk to cross-sectional
regression factor models, of interest to PMs. This paper is an abridged version.
21.
Nearest Neighbor Technique for a Positive Definite Correlation

Matrix in Advanced Stressed VAR
Number of pages: 7 Posted: 13 Jul 2016 Last Revised: 22 Nov 2016
• View PDF
• Download
•
Abstract:
We present a new technique for obtaining a positive definite (PD) correlation matrix from
a stressed target matrix within the context of Advanced Stressed Value at Risk, (cf. Dash
). The technique uses the spherical decomposition and a “nearest neighbor” technique.
The advantage is that control over the scenario specified by the risk manager for the
target matrix is achieved in that a maximal number of correlations specified by the target
matrix scenario is maintained.
22.
Risk Tails and General Orthonormal Polynomials

Jan Dash, Harvey J. Stein and Mario Bondioli
• View PDF
• Download
•
Abstract:
In order to characterize a statistical probability distribution p(x) of a variable x, the
moments of the distribution are used; the first two of which are the mean and standard
deviation. The z-score is often used to characterize data points of x (e.g. outliers with
large z-scores). Polynomials with respect to p(x) as the measure in the orthogonality
relation for the polynomials can be constructed. These generalize the ubiquitous z-score.
These polynomials (which we call GONPOMs) can be useful to refine the
characterization of data. Specifically they can be used in a targeted way to characterize
the change in shape of a distribution, e.g. for risk tails. It turns out that the GONPOMs
are (non-standard) polynomials first described by Chebyshev. Our purpose here is to
describe the theory and give a simple prototype numerical example.
23.
Stable Reduced-Noise 'Macro' SSA - Based Correlations for

Long-Term Counterparty Risk Management
Jan Dash, Xipei Yang, Harvey J. Stein and Mario Bondioli
• View PDF
• Download
•
Abstract:
We introduce a methodology from geophysics, Singular Spectrum Analysis (SSA), to
obtain stable, noise-cleaned correlations for long term risk (e.g. counterparty risk). SSA is
applied to time series to smooth them in a robust manner. The SSA-smoothed time series
are then used to obtain the correlations. We call these “macro” correlations because they
are determined with macroscopic time scales. Stable correlations are desirable to suppress
noise from short time scales that make risk measures unstable. If correlations move
around, risk measures also move around, making business decisions difficult. SSA-based
correlations ameliorate this business problem.
24.
MSSA vs. Multivariate Regularized Expectation Maximization

for Data Cleaning
Number of pages: 22 Posted: 20 Jul 2016
• View PDF
• Download
•
Abstract:
Previously we introduced Singular Spectrum Analysis SSA and its multivariate extension
MSSA as a powerful tool for cleaning data. Here we compare MSSA with the data filling
algorithm M-REM (Multivariate Regularized Expectation Maximization). We compare
theoretical methodology, numerical stability, algorithm capability, flexibility and speed.
Theory and numerical tests indicate that MSSA is superior to M-REM for data cleaning.
25.
Describing Crises with a Critical Exponent of the Reggeon Field

Theory
• View PDF
• Download
•
Abstract:
We present evidence that markets in crisis can be described by a critical exponent of the
nonlinear-diffusion Reggeon Field Theory, calculated 40 years ago, with no free
parameters, translated to finance. We propose this as a benchmark for average crisis
behavior, to which individual crises can be rich or cheap. In another paper we present a
quantitative model for the probability of equity crises in advance. An earlier paper
contained a summary.
26.
Non-Leading Eigenvalue Distributions, RMT, and Correlations

• View PDF
• Download
•
Abstract:
We showed that Singular Spectrum Analysis (SSA) applied to time series yields better
correlations for risk simulations. This involved comparing SSA-based correlations with
standard correlations and to noise, a zero correlation Wishart random matrix (WRM). We
complete this testing here. We also present tractable analytic approximate WRM results
that we used in the analysis: (1) leading and non-leading eigenvalue distributions of a
WRM, (2) eigenvalue spacing of WRMs, and (3) eigenvector components of WRMs.
27.
HYVAR (Hybrid VAR): HVAR Mixed with MC-HVAR

Number of pages: 8 Posted: 12 Jul 2016 Last Revised: 14 Feb 2017
• View PDF
• Download
•
Abstract:
We construct “Hybrid Value at Risk” (HYVAR) that is an arbitrary mixture of Historical
VAR and Monte Carlo VAR. The procedure is capable of retaining both the correlation
matrix of the original time series and also jumps/‘fat tails’. For this reason HYVAR
provides more realistic scenarios, and as many as desired. The main idea is to use a
“mixing angle” to mix HVAR with Monte Carlo HVAR.

DASH SSRN Finance/Risk Papers 2018

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

DASH SSRN Finance/Risk Papers 2018

Uploaded by

Copyright:

Available Formats

Jan Dash

Advanced Idiosyncratic Risk and Multi-Factor Models

Finally, we indicate the generalization to cross-sectional regression factor models.

In summary this work achieves two results:

Psychology, Stock/FX Trading, and Option Prices

Smart Monte Carlo, Path Integrals, and American Options

Market Crises, Earthquakes, and the Reggeon Field Theory

Cleaning Financial Data Using SSA and MSSA

Path Integrals and Greeks

Data Spike Cleaning with MSSA

The Macro-Micro Model, Trends vs. Noise, and SSA - I

SSA, Random Matrix Theory, and Noise-Reduced Correlations

Predicting Equity Crises, Critical Exponents, and Earthquakes -

Cleaning Data with Real-World Updating Using MSSA

Path Integrals and Smart Monte Carlo - I

Noise-Reduced Correlations, the Signal to Noise Ratio, and SSA

Analytic Solution to the Two Dimension Merton Model

Macro-Micro, Trends vs. Noise, and SSA - II

Path Integrals and Smart Monte Carlo - II

A Distressed Bond Model

Introduction to Noise-Reduced Correlations Using Singular

Speeding Up VAR with Smart Monte Carlo

Advanced Idiosyncratic Risk and Multi-Factor Models – Short

Nearest Neighbor Technique for a Positive Definite Correlation

Risk Tails and General Orthonormal Polynomials

Stable Reduced-Noise 'Macro' SSA - Based Correlations for

MSSA vs. Multivariate Regularized Expectation Maximization

Describing Crises with a Critical Exponent of the Reggeon Field

Non-Leading Eigenvalue Distributions, RMT, and Correlations

HYVAR (Hybrid VAR): HVAR Mixed with MC-HVAR

You might also like