Stock returns and Cash Flows: a new asset pricing approach

10 
Abstract (English version) 
Prior theories and models on assets pricing have highlighted the demand side of the security 
prices to evaluate the stocks’ performance.  However, fundamental analysis and accounting 
measurement theory have been considered ad hoc for another approach related to the intrinsic 
value of the stocks; value investors are focused on the main concept “price is what you pay but 
value is what you get” – (Warren Buffet, 2018). 
On this purpose, the dissertation is focused on a non-conventional profitability measure, at least 
in terms of assets pricing models, where dividends or profits are widely used. The attention is 
focused on a proxy measure of Operating Cash Flows: the “Ebitda after Capex”. 
The relantioship returns – cash flows’ volatililty has been examined throught an empirical 
analysis conducted on the stocks of the S&P500 Index combining the main quantitative and 
statistical approach with a qualitative overview respect the macroeconomic background. 
Starting from a correlation rolling window approach, three different regressions techniques 
have been implemented; the simple Ordinary Least Squares regressions (OLS), the linear 
Quantile (LQR) regression and the Multiple regression model (MLR), all  performed at 
different levels in terms of stocks (QoQ and YoY) and sectors (MoM, QoQ, YoY). 
The cross-sectional and time-series results support the effects of cash flow’ volatility on the 
stocks’ performance and highlighted its sensitivity respect not only the different short-term and 
long-term horizons, but also in terms of sector’ exposure.  
 
Keywords: Asset pricing, Volatility, Return, Quantile Regression, Stock pricing, Cash flow, 
Financial modelling, value investing, CAPM, Fama – French models.
12 
Introduction 
The motivations driving an investor to invest are led to the prospective of maximizing his 
wealth in order to get profits. In the asset’s allocation framework, stock’ returns are one of the 
most important factors that investors take into account; the higher the return is, the more the 
investor is attracted. The returns of most of the financial assets are directly derived from the 
price at which they are trading on the market, hence investors usually are focused on market 
behavior in order to find stocks with the more convenient risk-return payoff. However, this is 
not the only way. 
The passive investment management is based on considering the true value of the stock as 
proposed by the “Value investing” approach of Benjamin Graham and David Dodd, underlining 
that “price is what you pay but value is what you get”- (Warren Buffet, 2018). 
Trying to follow and apply this approach is not easy, as it might seem. First of all, because is 
not easy to define a measure or more in general a parameter able to express the “intrinsic value” 
of a stock. On this purpose, Fama and French, readjusting the CAPM model, proposed a more 
complex model where the “value stocks” could be identified considering the stocks with the 
higher Book-to-Market ratio. The three-factors model has changed over time, incorporating 
another characteristic: the profitability of the companies, useful to catch the profits’ margin of 
an investment. In reality, it has not been mentioned any particular formula to identify a “value 
stock”. Hence, it is crucial underline that even if the level of stock’ prices is mainly affected by 
market conditions, market performance and risk rate, the intrinsic value of a stock reflected on 
the market, can be expressed by the cash.  
On this purpose, the dissertation is focused on a non-conventional profitability measure, at least 
in terms of assets pricing models, where dividends or profits are widely used. The attention is 
focused on a proxy measure of Operating Cash Flows: the “Ebitda after Capex”.  More in detail,
13 
I extended the financial literature presenting an analysis that express the correlation among the 
cash flows of the companies of the S&P 500 Index and the performance of the stocks reflected 
in the returns, combining the main quantitative and statistical approach with a qualitative 
overview related to the macroeconomic background. 
The remainder part of the dissertation is organized as follows. After presenting the positive but 
also negative aspects related to the use of other measures of profitability, such as Dividends or 
Profits, the attention has been focused on the directly measure of cash flows. Cash flows are an 
accounting standard measure that, in addition of providing information on the firm's operating, 
investing and financing activities, provide a complete picture of the firm’ busyness life cycle, 
useful to evaluate the intrinsic value reflected on the stock’ price.   
In general, analysts and financial professionals are frequently referred to the terms EBITDA, 
Cash Flow (CF), Free Cash Flow (FCF), Free Cash Flow to Equity (FCFE), and Free Cash 
Flow to the Firm (FCFF or Unlevered Free Cash Flow). Among all these profitability measures, 
I pointed out the attention on Earnings before interest, taxes, depreciation, and amortization 
(EBITDA) as proxy measure of Cash from Operating activities (CFO). From an historical point 
of view, it can be noticed a higher attention not only on the concept of Ebitda as financial 
measure but also as a metric used for conducting investment decisions especially after the so-
called Leveraged Buy-Out (LBO) boom and the development of the Private Equities and 
Venture Capitals. From a strictly accounting point of view, analyzing and comparing all the 
positive aspects of the Ebitda respect the other measures, has been helpful to answer to one of 
the main criticisms pointed out by Warren Buffet: “Does management think the tooth fairy pays 
for capital expenditures?”.  
Hence for my analysis, I decided to not consider only the merely Ebitda metric’s but a more 
complex proxy measurement of cash flows that is the “Ebitda less Capex”, trying to follow the 
value approach, as investment strategy. In the core part of the dissertation the empirical analysis
14 
has been presented. Going into the deeper of the methodology, I firstly evaluated the stocks’ 
returns - cash flows’ volatility relationship, preforming a correlation based on a rolling window 
modeling approach. Subsequently, I constructed a synthetic ���������� ( ������ ,�������� )
 for which I used 
three different regressions techniques; the simple Ordinary Least Squares regressions (OLS), 
the linear Quantile (LQR) regression and the Multiple regression model (MLR). The whole 
story of the variability of the relationship cannot be obtained looking at just one-time frame, 
because yet over the longer time horizon adopted, the correlation varied. Hence, I stressed 
strategies and procedures performing empirical analyses at different levels; a Stock Market 
level analysis and a Sector level analysis with the aim of understanding why the relationship is 
stronger or weaker according to the different GICS classification. Every model has been 
reperformed considering different time horizons for the data (quarter on quarter, year on year 
and month over month). An insight has been presented analyzing not only the returns – cash 
flow’ volatility but also the effects of cash flows directly on stock prices. 
I concluded the dissertation underlining that while other studies aimed on trying to understand 
if the relationship among the returns – cash flow’ volatility is positive or negative, I investigated 
the time frame’ sensitivity and hence the variability of the cash flow’ volatility with respect to 
the short-term and long-term horizons. Finally, I tried to draw some conclusions also by looking 
forward at what can be done afterwards the empirical evidences reached and presented on the 
last section of the dissertation - (Conclusions: what is the next step?).
15 
 
Chapter 1 
 
Assets Pricing Models: Demand and 
Production side Approach 
 
Prior theories, researches and models on assets pricing have highlighted the demand side of the 
security prices to evaluate the performance of the stocks. With the demand side approach, the 
risk preferences of investors are considered in determining the security pricing. It is also called, 
consumption side since are the investors those who “consume” the assets. The demand, or 
consumption approach, has clearly been chosen as one of the most relevant aspects to consider 
for the estimation of the assets’ return. This is clear, looking at the literature background such 
as the Capital Assets Pricing Model (CAPM) and the Fama-French Factor models.  
 
On the other side, the production approach is based on considering and analyzing the company’ 
profitability aspects, usually the levels of Dividends or Profits, since they have an impact on 
the security prices.  In fact, investors use financial information, such as profitability and other 
measures related to cash, with the aim of evaluating a company’s intrinsic value and predict the 
future returns. However, using these aspects to evaluate the stock prices, is not possible every 
time, since there are many issues and problems to consider, presented in detail below1.  
 
1.1 Dividends: Shareholders’ Reward 
 
1 Benjamin A. Jansen: “Cash Flow Growth and Stock Return”.
16 
Dividends are usually used to reward shareholders since, by definition: they are a distribution 
of a part of the profits by a company to its shareholders. The interpretation of dividends, for 
many aspects, can be associated to the role of the profits. Dividends, in fact, can be analyzed in 
order to catch the ability, or not, of the company to earn profits constantly over time. Moreover, 
an increasing process in the distribution of dividends can be the signal for a good desirable 
future performance of the company2. 
Concerning dividends, there are increasingly fewer companies distributing them. The 
decreasing flow is not a recent event but has started time ago3.  
Going back in the history, just to contextualize what has been said so far and having a reference 
period for investigating the evolution of this process, we can say that in 1998 only 20.7% of the 
companies listed on the main US stocks markets (NYSE, AMEX and NASDQ) paid dividends, 
as noted by the Fama and French’s research published on 1999.  
4 
 
As stated before, the increasing process regarding the decreasing number of companies paying 
dividends, has started since 1926. In fact, over the time, the number of companies listed on the 
 
2William W. Priest, Steven D. Bleiberg, Michael A. Welhoelter: “Winning at Active Management: The Essential 
Roles of Culture, Philosophy and technology”.  
3 Benjamin A. Jansen: “Cash Flow Growth and Stock Return”. 
4 Pablo Fernandez: Valuation Methods and Shareholder Value Creation. Pg. 110. 
Table 1: Dividend Distribution by listed companies
17 
main US stock markets has increased and the share of companies that did not pay dividends has 
increased, respectively.  
In 1926-62 the total number of listed companies was 787 with a share of companies that were 
not pay dividends of 25.3%. The increasing process, both for the number of listed companies 
and for the number of companies that did not pay dividends, has reached its peak on the 1998 
where on 5.655 companies, only 20.7% of them paid dividends. Considering the entire time 
interval analyzed, the percentage of the companies that have not played dividends, increased of 
the 2,13%5 from the year 1926-62 until the year 1998.  
This reduction has not only caused by the fact that more and more companies decided to not 
pay dividends. From 81 companies that have never paid dividends, on the 1926-62, representing 
the 10.29%6 of the total, the incremental process has arrived at 3.981 on the 1998, representing 
the 70.39% 7 of the total. 
Apart from dividends, shares repurchase is one of the most similar ways for a company to 
distribute cash to its shareholders. It was noticed that stocks repurchase, the so-called buyback, 
have increased8. This is one of the events that can be useful to explain and understand why the 
dividends paying process has decreased. In fact, the buyback is an opposite process respect the 
payment of dividends, hence if the buyback of stocks increases, the dividends distributed 
decrease. 
 
Specifically, from 1980 to 1990, the aggregate value of stock repurchased on the New York 
Stock Exchange (NYSE), the American Stock Exchange (ASE), and the National Association 
 
5  (79.3%-25.3%)/ (25.3%) 
6 (81/787) x100 
7 (3981/5655) x100 
8 Pablo Fernandez: Valuation Methods and Shareholder Value Creation. Pg. 109-115.