Page 1 of 7
119
Archives of Business Review – Vol. 8, No.11
Publication Date: November 25, 2020
DOI: 10.14738/abr.811.9312.
Samsa, G. (2020). Why Does Momentum Persist? Archives of Business Research, 8(11). 119-125.
Why Does Momentum Persist?
Greg Samsa
Ph.D, Professor, Department of Biostatistics
and Bioinformatics, Durham NC, USA
ABSTRACT
The phenomenon of short-term momentum in intriguing because it
directly contradicts the notion that in an efficient market stock prices
should lack memory. Three classes of possible explanation for
momentum are (1) it is consistent with tenets of efficient markets after
appropriate risk adjustment; (2) it is consistent with tenets of
behavioral finance because of investors' cognitive biases; and (3) it is
consistent with structurally-based positive feedback loops. The
presence of extreme bubbles provides evidence against the first
explanation. The fact that prices are effectively set by institutional
investors who are aware of the cognitive biases in question and have a
financial incentive to avoid them provides evidence against the second.
Structurally-based explanations include the short-term incentives of
institutional money managers and the impact of indexing. We believe
that considering structural factors affecting the behavior of stock
prices provides an additional perspective, to be used in combination
with behavioral finance and market efficiency.
Keywords: behavioral finance; investment strategies; market efficiency;
momentum.
INTRODUCTION
The stock market can be described as a pas de deux between value and momentum, the latter of
which is of primary interest here. A key early citation in the literature on momentum investing is
by Jegadeesh and Titman [1]. For example, they created portfolios of stocks which had performed
particularly well during the previous 6-12 months (i.e., exhibiting positive price momentum) and
observed that such portfolios generated a significant excess return during the next such time
period.
The phenomenon of momentum in intriguing because it directly contradicts the notion of a fairly
weak form of capital market efficiency (i.e., that markets do not have memory with respect to past
prices). [2] Indeed, while the efficient market hypothesis approaches all putative market anomalies
with skepticism, it considers with particular skepticism the persistence of anomalies after they
have been publicized. [3] Proponents of behavioral finance assert that the likely cause of such
recurrent anomalies (if they actually exist) are cognitive biases inherent in human nature.
Our interest is not in every possible manifestation of momentum in the stock market. Instead, it is
in relatively large-capitalization stocks using an investment horizon of 3-12 months. More
specifically, we exclude stocks whose capitalization is so small that their prices are effectively set
Page 2 of 7
Samsa, G. (2020). Why Does Momentum Persist? Archives of Business Research, 8(11). 119-125.
URL: http://dx.doi.org/10.14738/abr.811.9312 120
by the behavior of individual investors, and instead focus on stocks whose prices depend upon the
behavior of their institutional counterparts (e.g., mutual funds, pension funds, hedge funds). The
rationale is that institutional investors are aware of cognitive biases and have a strong financial
incentive to avoid them. We exclude very short-term time horizons and thus, for example, treat
the literature on investors' immediate response to corporate news as outside the present scope.
We also exclude time periods which are sufficiently long that value tends to outweigh momentum,
as manifested by regression toward the mean [4], for example, 24 months and beyond. [1]
Assuming for the sake of argument that institutional investors can, if they choose, avoid the various
cognitive biases catalogued by behavioral finance, we ask what can instead cause momentum.
Although the literature is not entirely consistent about this, for the present purposes we stipulate
that (a) short-term momentum exists (indeed, this phenomenon has been observed across various
asset classes); and (b) momentum tends to reverse itself over longer time horizons. [5] Because of
the tendency for momentum to be reversed, we in effect are assuming that momentum has induced
a temporary mispricing.
OPERATIONAL DEFINITION OF MOMENTUM
Momentum can be operationally defined as the relative persistence of winning and losing
investments. In other words, investments which are performing well will tend to continue to do
so, at least in the short term, and investments which are performing poorly will continue to do so,
at least in the short term. The former will be termed "positive momentum" and the latter termed
"negative momentum". The operational definition of short-term depends on context: for a day
trader the context might be a matter of minutes, whereas for the present purposes this is 3-12
months.
SOME POSSIBLE CAUSES OF MOMENTUM
Subrahmanyam [2], in a review article which provides an excellent instruction to the literature on
momentum investing, lists some possible explanations for the apparent excess returns associated
with this strategy. One set of explanations (not considered in detail here, given that we have
assumed that momentum induces a temporary mispricing) is that momentum is nevertheless
consistent with efficient markets, for example:
• High past returns are correlated with high growth rate risk, which leads to higher required
returns in the future [6] (more generally, that proper risk adjustment will remove the
apparent short-term excess returns of momentum-based strategies).
• Uncertainty about the implication of news resolves slowly rather than immediately [7],
implying that what appears to be a single economic signal in fact is many.
• Another set of explanations is consistent with behavioral finance, namely that cognitive and
behavioral biases cause stocks to be mispriced, that these biases are corrected over a period
of 3-12 months, and that the manifestation of these corrections is momentum. Among the
causes of this putative mispricing include the following:
• Overconfidence causes investors to initially discount information which should be relevant
to stock price. [8]
• Investors underreact to news because of a psychological tendency to sell winners and retain
losers. [9]
• Investors do not process news simultaneously. [10]
Page 3 of 7
Archives of Business Research (ABR) Vol 8, Issue 11, November-2020
121
• Investors use overly simplistic mental models when evaluating stocks. [11]
• Investors underreact to information arriving in small bits. [12]
• Investors underreact to news which is inconsistent with their prior beliefs. [13]
• Investors overreact to noise in stock prices, and temporarily treat this noise as a continuing
signal. [14]
• Yet another set of explanations pertains to the role of market dynamics in determining stock
prices, for example:
• Stock prices tend to show greater covariance than do corporate fundamentals (i.e., they
"move together more than they should"). [15]
• The incentives of institutional investors favor momentum.
• When investors apply momentum-based trading strategies this imparts stock prices with
additional momentum.
SIGNAL-BASED CAUSES OF MOMENTUM
The above behavioral science explanations posit a temporary mispricing in response to an
improperly interpreted economic signal. For example, suppose that the stock in question has
outperformed the market during the previous 6 months, and thus that the economic "signal" is
(unexpectedly) positive information about that company revealed during that time period. For
most of these explanations investors initially underreact to this economic signal, causing a
temporary mispricing which is then corrected during the next 6 months by positive momentum.
A counter-argument raises three questions. First, aren't the people whose trading behavior
effectively set stock prices highly-trained analysts who are aware of the above putative biases?
After all, institutional investors aren't the 30 sophomore economics students, volunteers recruited
from a local shopping mall, etc., sometimes used to demonstrate the behavioral biases in question.
Second, don't these same institutional investors have a strong financial incentive to override these
biases (i.e., by taking advantage of this temporary mispricing)? Finally, why aren't the prices of
"momentum stocks" immediately bid up to reflect their anticipated performance, thus correcting
the reaction to the signal (and, in the process, eliminating the short-term correlation among stock
prices)?
STRUCTURAL CAUSES OF MOMENTUM
The market-dynamics-based explanations above each propose some form of positive feedback loop
embedded within the structure of the market -- indeed, a feedback loop which is sufficiently
powerful to override the usual economic forces acting to cause stock prices to immediately account
for all available information. This positive feedback could be initially triggered by previous price
changes, by changes in the prices of related stocks, or, indeed, a feedback loop need not have an
initial "cause" beyond random noise. In general, a feedback loop need not necessarily imply a
mispricing, although in this case we assume that it is sufficiently strong as to overshoot true value,
inducing underperformance in the longer term.
An extreme example of a positive feedback loop is a bubble whereby stock prices are temporarily
bid up beyond any reasonable economic justification. Such bubbles, ranging from the tulip bulb
mania in the 1600s to recent bubbles in cannabis and crypto-currencies, provides some of the
strongest evidence against the hypothesis that markets are always efficient and, indeed, leaves
Page 4 of 7
Samsa, G. (2020). Why Does Momentum Persist? Archives of Business Research, 8(11). 119-125.
URL: http://dx.doi.org/10.14738/abr.811.9312 122
believers in efficient markets with the somewhat unsatisfying response that "It's one thing to
recognize a bubble, but since you don't know when it will burst it's quite hard to profit from this
recognition, and an anomality which can't be easily exploited doesn't really count".
A number of factors could be sufficiently strong to override the economic forces which typically
cause stock prices to account for all available information. One such factor pertains to the
incentives of mutual fund managers, hedge fund managers, and others who trade regularly and
thus whose actions tend to set prices in the short-term. For them, assets under management is a
key component of compensation. Reports are made quarterly, and customers tend to flee from
funds with poor short-term performance. Thus, there is little incentive to create a portfolio with
excellent long-term prospects if the manager won't be around to enjoy its benefits. Among others,
this incentive can induce "window dressing", where portfolio managers sell losers and buy winners
before a quarterly report of portfolio holdings, and "herding", which allows managers to contend
that their performance is no worse than others, even if it is poor in absolute terms.
Eventually, these "perverse incentives" are overridden by the actions of the largest investors such
as pension fund managers. This latter group of investors can't trade regularly, because of the
associated friction costs, and by necessity must move at a pace which recalls a Monte Python skit
about low-energy cricket being played by various items of furniture. Their incentive is to slowly
accumulate shares of sound businesses at advantageous prices, which purpose is aligned with
investment rather than speculation, and they eventually bring the "value" to the dance between
value and momentum.
Another structural factor is correlation of returns within funds. For example, a fund which suffers
poor returns will have "outflows", requiring the manager to sell shares, causing the stocks within
that portfolio to perform poorly. The impact is indiscriminate, and is exacerbated by herding when
different managers sell the same stocks.
Yet another structural factor is the increasing popularity of indexing. Stocks which are performing
well receive greater weight as indices are rebalanced, and this induces positive momentum.
Indeed, the more the amount of indexing the greater is the impact on momentum. Even if the
incentives of regular traders were changed, the impact of indexing would remain.
Because of these structural factors, it is reasonable to anticipate that the impact of momentum will
continue into the foreseeable future.
THE ENDGAME
One can ask how all of this will end. Our answer is "asymmetrically". For clarity of exposition we
will consider two bubbles, one positive and one negative. Essentially the same argument will apply
to less dramatic examples of feedback loops.
A current example of a negative feedback loop is that of energy stocks. Without recapitulating the
multiple reasons why such stocks are profoundly unpopular at present, we note that corporate
management has the option of sending three concrete signals to highlight their belief that their
stock price has fallen below true economic value. Management can buy back shares at
(presumably) bargain prices, thus increasing earnings per share, which in turn should eventually
Page 6 of 7
Samsa, G. (2020). Why Does Momentum Persist? Archives of Business Research, 8(11). 119-125.
URL: http://dx.doi.org/10.14738/abr.811.9312 124
momentum stocks tend to be overpriced, and buying them will reduce my long-term returns, but I
anticipate that my short-term returns will be above average, and also that I can beat my
competitors out the door when the time comes".
Similarly, interpreting momentum purely from the perspective of behavioral finance induces
problems of its own -- most particularly, how to address the fact that institutional investors "know
better". Indeed, we argue that institutional investors are not generally driven by an "invisible hand
of cognitive biases", but rather that they have an incentive to behave as if they suffer from the biases
in question.
We have focused on stocks with such large capitalizations that their prices are driven by the action
of professional institutional investors rather than individual amateur investors, as we have posited
that the actions of the former are likely to be more "rational" than the actions of the latter.
However, the distinction between these two groups is blurred by the impact of advertising. More
specifically, when institutions advertise momentum funds based on short-term performance and
the exciting prospects of the innovative companies which their portfolios contain they in effect are
encouraging their less-well-informed customers to act upon various cognitive biases, which in turn
provides additional incentives for fund managers to behave as if they also suffer from these same
biases.
We believe that considering structural incentives affecting the behavior of stock prices provides an
additional perspective which is helpful. As above, this perspective should not be applied in
isolation, but instead as a supplement to notions from behavioral finance and market efficiency.
References
[1] Jegadeesh N, Titman S. 1993. Returns to buying winners and selling losers: implications for stock market
efficiency. Journal of Finance. 48, 65-91.
[2] Subrahmanyam A. 2018. Equity market momentum: a synthesis of the literature and suggestions for future
work. Pacific-Basin Finance Journal. 51:291-296. https://doi.org/10.1016/j.pacfiin.2018.08.004.
[3] McLean D, Pontiff J. 2016. Does academic research destroy stock market predictability? Journal of Finance. 71,5-
32.
[4] Samsa G. 2014. A new critique of the traditional method for empirically estimating the returns associated with
strategies of stock investment: impact of fixing the investment horion. Bentley University Case Studies in
Business Industry and Government Statisitis. 5(2): 102-110.
[5] Asness C, Moskowitz T, Pedersen L. 2013. Value and momentum everywhere. Journal of Finance. 68, 929-985.
[6] Johnson TC. 2002. Rational momentum effects. Journal of Finance, 57; 585-608.
[7] Holden CW, Subrahmanyam A. 2002. News events, information acquisition, and serial correlation. Journal of
Business. 75; 1-32.
[8] Daniel K, Moskowitz T. Momentum crashes. Journal of Financial Economics. 122, 21-247.
[9] Grinblatt M, Han B. 2005. Prospect theory, mental accounting, and momentum. Journal of Financial Economics.
78, 311-339.
[10] Hong H, Stein J. 1999. A unified theory of underreaction, momentum trading and overreaction in asset markets.
Journal of Finance. 54, 2143-2184.
[11] Hong H, Stein J, Yu J. 2007. Simple forecasts and paradigm shifts. Journal of Finance. 62, 1207-1242.