“May the best of your past be the worst of your future” Tony Evans Many payoffs that exist today are based upon the performance of multiple assets. When an option derives its value from the price of multiple assets, the relationships between these assets become important. Best-of options are an example of multi-asset options. They are based on the distribution of the best asset performer. The relationships between the underlyings are defined using correlations that gives us the strength and direction of a linear relationship between different underlyings. Correlation is an extremely important concept for multi-asset financial contracts because, unlike single-asset options, their value does not only depend on the implied volatilities of the underlying asset but also on the correlations between these assets. We are taking into account its fundamental role by analyzing the distribution of the maximum of undelyings, under different correlation values. The analysis starts considering Gaussian undelyings, propaedeutic assumption inspired by Bachelier model. We consider first a basket of 2 underlying, generalizing later to N underlyings. We hence obtain the density distribution and moments for a generic case. We then apply a moment matching procedure to compare the analytic moments with the empirical ones, recognizing the nearest Gaussian distribution. In order to better adhere to reality, we assume that the dynamics are lognormal with constant volatility, according to Black-Scholes-Merton model. Then, the existence of volatility smile entails the need to move to a more sophisticated model that would be able to describe this pattern, assuming different values of volatility to price options at different strikes. Having hence selected the SABR local volatility model, we analyze the presence of implied correlation skews generated by an equally weighted basket option, that we find out to be consistent with the empirical evidence in the market.
“May the best of your past be the worst of your future” Tony Evans Molti payoffs che esistono oggi si basano sulla performance di più assets. Quando un’opzione deriva il suo valore dal prezzo di più assets, le relazioni tra questi diventano importanti. Le opzioni Best-of sono un esempio di opzioni multi-asset. Si basano sulla distribuzione del migliore performer tra gli assets. Le relazioni tra i sottostanti sono definite utilizzando le correlazioni, che forniscono il grado e la direzione di una relazione lineare tra i differenti sottostanti. La correlazione è un concetto estremamente importante perché, a differenza delle opzioni single-asset, il valore di un’opzione multi-asset non dipende solo dalle volatilità implicite dell’asset sottostante ma anche dalle correlazioni tra questi. Prendiamo in considerazione il ruolo fondamentale della correlazione analizzando la distribuzione del massimo di undelyngs, sotto diversi valori di correlazione. L’analisi comincia con undelyings Gaussiani, assunzione propedeutica ispirata al modello Bachelier. Consideriamo prima un basket di 2 sottostanti, generalizzando poi a N sottostanti. Otteniamo quindi la distribuzione di densità e i momenti per un caso generico. Applichiamo poi il metodo dei momenti per confrontare i momenti analitici con quelli empirici, individuando la distribuzione gaussiana più vicina. Per aderire meglio alla realtà, introduciamo un grado aggiuntivo di complicazione, assumendo che la dinamica sia lognormale con volatilità costante, secondo il modello Black-Scholes-Merton. Quindi, l’esistenza dello smile di volatilità comporta la necessità di passare a un modello più sofisticato che sia in grado di descrivere questo andamento, assumendo diversi valori di volatilità per valutare le opzioni a diversi strike. Avendo selezionato il modello di volatilità locale SABR, analizziamo la presenza di skew di correlazioni implicite generate da un basket equamente pesato, che scopriamo essere coerenti con l’evidenza empirica del mercato.
Distribution of maximum of underlyings in BlackandScholes and local volatility frameworks
PENNISI, SILVIA
2020/2021
Abstract
“May the best of your past be the worst of your future” Tony Evans Many payoffs that exist today are based upon the performance of multiple assets. When an option derives its value from the price of multiple assets, the relationships between these assets become important. Best-of options are an example of multi-asset options. They are based on the distribution of the best asset performer. The relationships between the underlyings are defined using correlations that gives us the strength and direction of a linear relationship between different underlyings. Correlation is an extremely important concept for multi-asset financial contracts because, unlike single-asset options, their value does not only depend on the implied volatilities of the underlying asset but also on the correlations between these assets. We are taking into account its fundamental role by analyzing the distribution of the maximum of undelyings, under different correlation values. The analysis starts considering Gaussian undelyings, propaedeutic assumption inspired by Bachelier model. We consider first a basket of 2 underlying, generalizing later to N underlyings. We hence obtain the density distribution and moments for a generic case. We then apply a moment matching procedure to compare the analytic moments with the empirical ones, recognizing the nearest Gaussian distribution. In order to better adhere to reality, we assume that the dynamics are lognormal with constant volatility, according to Black-Scholes-Merton model. Then, the existence of volatility smile entails the need to move to a more sophisticated model that would be able to describe this pattern, assuming different values of volatility to price options at different strikes. Having hence selected the SABR local volatility model, we analyze the presence of implied correlation skews generated by an equally weighted basket option, that we find out to be consistent with the empirical evidence in the market.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi_SilviaPennisi.pdf
non accessibile
Dimensione
4 MB
Formato
Adobe PDF
|
4 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/185805