Structured Retail Products Methodology and Analytics


One of Fairmat Cloud’s goals is to allow investors to compare financial products in an objective way and to provide insights (or intelligence information) on the likely future behaviour of the structured products we cover.

It’s a known fact that financial markets contain most of the information about the risk of the underlying factors driving the pay-outs: traded assets, for example prices of call/put options on indices provides information about the riskiness of the underlying indices. This insight is one of the key ingredients of our products comparison framework.

In the case of the structured products there are additional complications: products performances are non-linearly dependent with respect to the evolution of underlying assets. By exploiting the detailed relationships between underlying drivers and products pay-out which can be derived by the products’ term sheets, we build accurate simulation models of the pay-outs in order to take non linearities into account.

Finally our independent products sheets contain dynamic information calculated using observable market data and as much as possible they are market consistent, meaning that they are obtained by calibrating theoretical models for the underlying assets’ evolution in terms of the observed market prices of primary assets which are determined by the equilibrium between supply and demand (thus reflecting traders’ s views).

All the offered analytics are calculated using public available quantitative information. We provide market consistent forward-looking analytics like early redemption probabilities, tail risks, etc. and historical analyses, in which we back-test the tested investment over the market history.

Please read below for further details about the methodology employed on each section of our independent product sheets.

Fair Value

Given the complexity and heterogeneity of the certificates object of our analyses, we employ a decomposition approach in which the structure of the contingent cash-flows and of the events triggering early redemption are expressed in terms of the considered risk factors. The risk factors evolution is then simulated using a Monte Carlo approach following the frameworks proposed by Boyle, Phelim P. in Options, a Monte Carlo Approach (1977) and Longstaff, Schwartz in Valuing American Options by Simulation: A Simple Least-Squares Approach (2001).

In order to simulate the evolution of the equity assets that underlie many of these certificates, we mainly use the Dupire local volatility model which is an industry standard for equity derivatives.

We model potential Issuers defaults using a reduced form intensity framework suitable for being calibrated to Credit Default Swap (CDS) data. For more details on this approach see Brigo, Mercurio - Interest Rate Models, Theory and Practice (2006).

Products Ranking

Ranking is calculated by ordering products by their score, a performance indicator which take into account expected return and risk calculated on an average growth scenario and for an average risk averse investor.

Figure 1: Example of product ranking, probability of outperforming domestic bond and fair value.

Risk-Return Analysis and Risk-Return Map

According to the risk-return trade-off which states that investments can render higher profits only if they are subjected to a greater possibility of being lost. In this analysis, we report the product's annualized expected returns vs the semi-deviation of the investment at maturity.

In order to do that, we simulate the capitalized product pay-out for several risk factors scenarios, and from those scenarios we calculate the empirical expected returns, and the semi deviation (volatility of negative events).

Furthermore, we offer a comparison of the product risk-return trade-off with respect to other structured products issued in the same market or currency.

Returns for a given market date are calculated with respect to the products' theoretical product prices (at the same market data) which are calculated by derivative pricing techniques.

Figure 2: Example of Risk-Reward trade-off, in which the current product is highlighted.

Comparison with the bond market

Aside comparing structured products which each other, we aim to compare the structured products with another popular investment asset class: the domestic sovereign bond. For any structured product we design a synthetic bond paying semi-annual coupons having the same maturity of the product.

For this analytics we use the probabilistic approach proposed in Minenna – A Quantitative Framework to Assess the Risk-Reward Profile of Non-Equity Products (2011) in which we compare the total return at maturity (in which early redemption pay-outs are compounded using a stochastic interest rate).

The expected returns of the product and the corresponding probability of occurrence are estimated in the cases where the certificate is found to be better, in line or worse as compared with the alternative risk-free investment. The circles displayed in the traffic light figure measures the impact to the investor of a given scenario and have an area proportional to the absolute value of the product returns in a given scenario times the scenario's probability.

Figure 3: Example of investment comparison with a domestic bond.

Tail Risk

Tail risk is calculated as the CVaR (expected shortfall) of product returns conditional to the occurrence of events with probability 1% and 5% and include credit risk.

Figure 4: Example of tail expected returns.

Early Redemption and Barriers

When the product has early redemption features, we calculate the probability of early redemption before the natural Maturity of the Certificate. This provide investors information about the expected duration of the investment.

Figure 5: Example of early redemption and capital protection loss probabilities calculation.

Outlook Scenarios

All the analysis reported above are also used in our screening system (ranking tool and risk-map) and are available on 5 different average growth (risk premium) scenarios. For any growth scenarios pricing is performed using several thousand Monte Carlo simulations.