# MASSIMO GUIDOLIN

## Topics for MSc. dissertations (research theses and not)

If you care for any of these topics, come and visit me during office hours to discuss whether we may have a match of incentives and goals.

No e-mails, personal visit is required.

Remember: no faculty member is forced to accept your as a supervisee, especially if this faculty member supervises already a dozen theses or more. Therefore, DROP BY AND DISCUSS topics, do not simply start working on them on your own.

IMPORTANT: EFFECTIVE JUNE 21, 2018, I WILL ACCEPT

__RESEARCH__THESIS IF AND ONLY IF AS STUDENT HAS SUCCESSFULLY WRITTEN MY OPTIONAL EXAM, 20541 - ADVANCED QUANTITATIVE METHODS FOR ASSET PRICING AND STRUCTURING, see
http://didattica.unibocconi.it/ts/tsn_anteprima2006.php?cod_ins=20541&anno=2018&IdPag=6066

Note that up to 6 optional exams can be taken (and will count for a student's final GPA).

__: I am looking for students willing to write a normal thesis (max 5 points), with a strong survey/review flavor (i.e., no data work or limited empirical work) on the following two topics:__

**News**
A. Machine learning and big data approaches applied to derivative pricing.

B. Machine learning and big data approaches applied to market risk measurement and portfolio return density forecasting.

E-mail me in case you are interested.

1. Asset pricing models applied to commodity returns.

Are commodities just another financial asset class or are they real assets? What is the practical and theoretical difference between a commodity and the futures written on them? How can commodities change the performance of a diversified portfolio? How can we explain the time-varying correlations between commodities and other asset classes and within the commodity class?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3225611

2. Asset pricing model uncertainty in asset management.

There is ample evidence that in financial economics we lack a generally accepted model for the mechanism generating asset returns. The uncertainty concerns both the nature of the factors/predictors and the mathematical form of such a model. In the meantime, plenty of applications of forecasting to finance have confirmed that it may be best to pool alternative models -- by forming forecast combinations -- than in fact selecting the best model. Alternatively, one may wonder whether the dynamic, regime shifts across different models may be modelled and fruitfully predicted. This project asks whether such combinations and or/regime shifting model selection framework may deliver in asset management. The risk adjusted performance should be assed through a process of portfolio selection.

3. Is Black-Scholes really dead?

In the very courses that you have been taught there is often a presumption -- explicit or implicit -- that in practice derivatives traders and users for hedging purposes would mostly use Black-Scholes model, because "this is all that is needed". Is this true?

In fact, a few provocative papers have even doubted that Black-Scholes has ever been really used by option traders at all.

There is little doubt however that at least since the early 2000s more complex models that take -although sometimes only heuristically and through empirical adaptation of Black-Scholes-- into account stochastic volatility and jumps and that are gradually but inexorably supplanting Black-Scholes also in practice.

One wonders whether, after all, this is justified at all:

_ does Black-Scholes allows us to forecast future realized volatility, through implicit volatility, better than other models do?

_ does Black-Scholes allows us to forecast future option prices better than other models do and therefore take profitable positions?

_ does Black-Scholes allows us to hedge risks more effectively than other models do, through the calculations of its simple greeks?

The goal of the thesis is to use state-of-the art statistical methods to test these hypotheses.

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=264655

4. Are Exchange Rates Really Unpredictable?

One of the oldest and plainest views in financial economics is that exchange rates are unpredictable and currency markets the most efficient of the world. Under some conditions, this is equivalent to state that the forecasts from a random walk with drift cannot be outperformed -- if not by chance -- by most (say, 95%) of the other models.

The goal of the thesis is to assess whether and how this traditional claim is correct in the light of:

_ modern statistical methods to test equal predictive accuracy among models,

_ state-of-the-art model, including no arbitrage one to elicit such forecasts.

5. The use of volatility and volatility products as an asset class.

Direct exposure to volatility has been made easier for a wide range of underlyings by the creation of standardized trading instruments, such as volatility index futures, variance swaps, and exchange-traded notes. In addition to short-term trading ideas, some investors have been seeking structural exposure to volatility because they consider it either a well-identified asset class or, at the very least, a set of strategies with strong diversifying potential for their portfolios. The remarkably strong negative correlation between implied volatility and equity prices during market downturns offers timely protection against the risk of capital loss. However, this insight is under debate: some papers report that although the classic diversification benefits are limited, modest allocations to short volatility exposure could have enhanced long-term returns, in one case increasing the portfolio’s combined Sharpe ratio. Who is right, who is wrong?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3076929

6. Hedge funds -- do they really create economic value to investors?

Do hedge funds really generate (strong) abnormal performance as commonly believed, or do they just expose investors to new, not well appreciated risks? Or in short: do hedge funds really hedge risks?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3225600

7. Smart beta strategies in asset management.

Recently, both academics and asset management practioners have realized that by:

_ tilting portfolio weights away from standard market capitalization, and/or

_ selecting portfolios that include assets that belong only to extreme deciles or terciles of the asset menu when sorted according to specifica criteria (e.g., the size of companies, the book-to-market or earnings-price ratio of companies, past realized variance, beta or momentum of stocks, etc.)

one can obtain strategies -- that have monicked as "smart beta" strategies -- that outperform the cap-weighted market portfolio, yield lower volatility, higher return-to-risk ratios, and (often positive alphas. Although there are many stories for why this could be the case, it is not really clear precisely how and why this occurs.

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2961653

When do credit default swap spreads go out-of-line vs. bond spreads and why? Is that phenomenon predictable? Is credit risk (or the opposite, a convenience yield for carrying inventory of special kinds of bonds) priced equally in the credit default swap (CDS) and bond markets?

9. The economic value of analysts forecasts and recommendations.

Financial analysts represent a strange and fascinating socieyt of which we still understand little in terms of their choices concerning forecast accuracy, strategic bias manipulation, relationship to stock recommendations, and more generally resolution of internal conflicts within general-purpose brokerage houses in which also corporate finance deals are developed. There is much to better understand...

10. Network Graphical Models in Finance.

Although markets are the usual forum where economic phenomena take place, many social and economic behaviors are not only mediated by prices, in the sense that it matters how agents are in contact with each other. Financial networks are natural examples: the main conduit is the intervening role of "connections": who is in direct or indirect contact with whom. This structure defines (and is possibly defined by) how information, prices and quantities reverberate in a particular system. Several studies have shown remarkable similarities between different large-scale networks that arise when humans interact, like friendship networks, networks of co-authorship and networks of e-mail correspondence) and stock

markets. Specifically, these networks tend to be sparse (the number of connections between nodes are of the same order as number of nodes, where in our networks the nodes represent individuals), they have small effective diameter (the so-called small world property) and power laws govern their degree distributions (i.e., the distribution of the number of connections associated with a specific node is power law distributed). How can network-based empirical and modelling approaches change our perception of financial phenomena, such as asset pricing, portfolio decisions, and systemic risk?

11. Financial econonometrics and portfolio selection perspectives on cryptocurrencies.

There is a been a considerable amount of chatter on the benefits and costs of cryptocurrencies that represent means of payment exploiting Distributed Ledger Technologies (blockchain). A recent literature has deployed modern techniques in time series econometrics and applied portfolio management to understand whether cryptocurrencies are subject to bubbles and may represent a novel, attractive asset class. This poses a number of questions:

_ Do cryptocurrencies contain bubble and their value records a disconnect from fundamental values?

_ How do cryptocurrencies and their volatilities and correlations move over time and does this affect their attractiveness as an asset class?

_ What are the benefits and risks investors face when they allocate wealth to cryptocurrencies?

12. Modelling and forecasting (using exogeneous variables as well as time series methods) the dynamics of the implied volatility surface for options.

The implied volatility surface transforms a panel of option prices into the (instantaneous or average) volatility under a given pricing model, for instance, Black and Scholes'. In this special but popular case, BS implied volatilities ought to be independent of both the strike (moneyness) of the contracts and of their time-to-expiration. In practice, they appear to strongly depend on both moneyness and maturity, describing rather typical shapes, such as smiles and skews, and generally a surface. Is the shape of the IV surface (IVS) predictable, in either a statistical or economic sense? Can it be useful to impose the absence arbitrage when models are built or used in forecasting? Can the dynamics of the IVS be linked to macroeconomic factors that are likely to enter in the SDF, the key pricing measures for all assets and portfolios?What are the key differences between the IVS dynamics for individual equity options and the equity index IVS?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=406697

13. Implied GARCH models in option prices and their use.

GARCH models represent the benchmark in the forecasting volatility. However, the models are typically specified under the physical measure, i.e., they are fit on realized asset returns data. Recently, GARCH models have been shown to also imply realistic option prices, under the risk neutral measure. Hence the idea to estimate GARCH models from option prices (hence under the risk neutral measure) to either forecast volatility or to base portfolio decisions on the forecasts so derives (with or without adjustments to go back to the physical measure). Can a GARCH model estimated from asset return data better assist portfolio decisions vs. a option-implied GARCH under the risk neutral measure? If so, with or without adjustment for risk premia? Is the conclusion affected by the specific type of GARCH model considered?

14. Comparing the growth and structure of the European ABS markets vs. the US one.

Many have argued that the roots of the 2007-2009 financial crisis have to be seeked in the enormous development of asset-backed security markets worldwide, especially in the US. In fact, the financial crises have marked a stop in the growth of such markets, that have instead regressed almost drying completely up. However, one legitimately asks whether such cycles of increasing and decreasing growth speed have occured already before and -- more importantly -- what economic factors may contribute to such dynamics over time and acrss types of assets. What is the effect of the strengthening of the regulations on such dynamics? What is the cost in terms of reduced ability to spread and transforms risks throughout the financial system as a result of the declining speed of growth of the asset-backed securities market?

15. Empirical corporate bond pricing.

What are the key features of the corporate bond market that -- for instance in terms of their pricing patterns or liquidity -- makes them different from other asset classes, and significantly Treasury and default-free riskless bonds?

16. The use of derivative and structured products in portfolio management.

Assuming that investors should buy structured, option strategy-like securities, what is their optimal share for a rational, optimizing portfolio? Moreover, when such a strategy gets actually implemented over time, do backtesting exercises reveal large risk-adjusted benefits?

17. Bubble in real estate and REITs.

Real estate, and in particular, residential housing markets have been accused to have represented the epicenter of the 2007-2009 financial crisis and not only in the United States. This would have resulted from a real estate bubble suddenly but predictably bursting during 2007. Other famous financial crises and resulting economic slumps have resulted from real estate bubbles gone sour. Are real estate markets really that prone to bubble and why? Could bubbles have been detected in real time?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2255943

18. Ambiguity in asset pricing and portfolio choice.

What if uncertainty were not risk, i.e., it were not objectively quantifiable? In this case we speak of Knightian uncertainty or more generally of ambiguity. There is ample empirical evidence and a common-sense presumption that investors will dislike ambiguity and be averse to it, to the point of demand a compensation (a risk premium) for bearing uncertainty affecting the outcomes of their financial decisions. Here the question is: how does ambiguity change what we commonly think about asset pricing and portfolio decisions?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1673494

19. Higher order moments in portfolio choice.

Although expected utility maximization represents a logically important benchmark in modern asset allocation models, a considerable portion of theory and practice of portfolio decisions can being cast in a simplistic mean-variance framework. Yet, it is well-recognized that investors care for much more than mean and variance of their portfolio returns, and in particular for its skewness and kurtosis. How are portfolio decisions affected by higher-order moments? Which are the most important and accurately estimable moments? Does the econometric model generating the moments matter for the realized portfolio performance and why?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=406720

20. Derivatives markets and unconventional monetary policies.

There is ample evidence that unconventional monetary policies enacted by the FED, the Bank of England, and the ECB between 2008 and 2012 (and still ongoing to some extent) have affected all sorts of financial markets and prices. Does this evidence extend to the markets for derivatives, and in particular options, given their forward-looking nature? And if so, why and how? Have unconventional monetary policies removed uncertainty that would otherwise plague financial markets thus altering the value of classical derivatives?

21. No arbitrage condition violations and/or misspricing in the structured derivatives market.

Why do retail investors and pensioners ever buy complex option portfolios that would give you a hard time in most Derivatives 2 exams? Are they correctly or even sensibly priced?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=430580

22. The economics of wars and conflicts.

There is an expanding literature on the ecomomic effects and causes of internal and external armed conflicts, in particular civil wars. Empirical work finds that low per capita incomes and slow economic growth are both robustly linked to civil war. Yet there is little consensus on the most effective policies to avert conflicts or promote postwar recovery. Moreover, there is a consensus that duing the 20th century, war and totalitarianism produced more famine deaths than did overpopulation and economic backwardness. Can we get to know more about these links?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=825889

23. Sentiment and social media in asset pricing and portfolio choice.

The evidence that human sentiment has a crucial effect on decision making induces the reasonable assumption that investors are also subject to this influence in financial markets and do not persistently behave rationally. Having said this, De Long et al. (1990) and Shu (2010) find that sentiment affects financial markets,and thus may lead to increased market fluctuation and excess volatility. Accordingly, sentiment must be considered as an integral part of the asset pricing theory and portfolio choice. However, the big issue remains: how should we measure sentiment?

1. Asset pricing models applied to commodity returns.

Are commodities just another financial asset class or are they real assets? What is the practical and theoretical difference between a commodity and the futures written on them? How can commodities change the performance of a diversified portfolio? How can we explain the time-varying correlations between commodities and other asset classes and within the commodity class?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3225611

2. Asset pricing model uncertainty in asset management.

There is ample evidence that in financial economics we lack a generally accepted model for the mechanism generating asset returns. The uncertainty concerns both the nature of the factors/predictors and the mathematical form of such a model. In the meantime, plenty of applications of forecasting to finance have confirmed that it may be best to pool alternative models -- by forming forecast combinations -- than in fact selecting the best model. Alternatively, one may wonder whether the dynamic, regime shifts across different models may be modelled and fruitfully predicted. This project asks whether such combinations and or/regime shifting model selection framework may deliver in asset management. The risk adjusted performance should be assed through a process of portfolio selection.

3. Is Black-Scholes really dead?

In the very courses that you have been taught there is often a presumption -- explicit or implicit -- that in practice derivatives traders and users for hedging purposes would mostly use Black-Scholes model, because "this is all that is needed". Is this true?

In fact, a few provocative papers have even doubted that Black-Scholes has ever been really used by option traders at all.

There is little doubt however that at least since the early 2000s more complex models that take -although sometimes only heuristically and through empirical adaptation of Black-Scholes-- into account stochastic volatility and jumps and that are gradually but inexorably supplanting Black-Scholes also in practice.

One wonders whether, after all, this is justified at all:

_ does Black-Scholes allows us to forecast future realized volatility, through implicit volatility, better than other models do?

_ does Black-Scholes allows us to forecast future option prices better than other models do and therefore take profitable positions?

_ does Black-Scholes allows us to hedge risks more effectively than other models do, through the calculations of its simple greeks?

The goal of the thesis is to use state-of-the art statistical methods to test these hypotheses.

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=264655

4. Are Exchange Rates Really Unpredictable?

One of the oldest and plainest views in financial economics is that exchange rates are unpredictable and currency markets the most efficient of the world. Under some conditions, this is equivalent to state that the forecasts from a random walk with drift cannot be outperformed -- if not by chance -- by most (say, 95%) of the other models.

The goal of the thesis is to assess whether and how this traditional claim is correct in the light of:

_ modern statistical methods to test equal predictive accuracy among models,

_ state-of-the-art model, including no arbitrage one to elicit such forecasts.

5. The use of volatility and volatility products as an asset class.

Direct exposure to volatility has been made easier for a wide range of underlyings by the creation of standardized trading instruments, such as volatility index futures, variance swaps, and exchange-traded notes. In addition to short-term trading ideas, some investors have been seeking structural exposure to volatility because they consider it either a well-identified asset class or, at the very least, a set of strategies with strong diversifying potential for their portfolios. The remarkably strong negative correlation between implied volatility and equity prices during market downturns offers timely protection against the risk of capital loss. However, this insight is under debate: some papers report that although the classic diversification benefits are limited, modest allocations to short volatility exposure could have enhanced long-term returns, in one case increasing the portfolio’s combined Sharpe ratio. Who is right, who is wrong?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3076929

6. Hedge funds -- do they really create economic value to investors?

Do hedge funds really generate (strong) abnormal performance as commonly believed, or do they just expose investors to new, not well appreciated risks? Or in short: do hedge funds really hedge risks?

See one example at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3225600

7. Smart beta stra

Last change 02/11/2018