Simple value and momentum strategies often end up with opposite market positions. One strategy succeeds when the other fails. There are two plausible reasons for this. First, value investors regularly bet against market trends that appear to ‘have gone too far’ by standard valuation metrics. Second, value stocks carry particularly high market risk or ‘bad beta’ and thus fare well when market risk premia are high and the market turns for the better. This typically coincides with ‘momentum crashes’ in oversold markets. As a consequence, value and momentum signals may be complementary. In particular, value strategies are not very profitable in normal times or bull markets but have produced extraordinary profits when being set up in the mature state of a bear market. Similarly, momentum signals can be adjusted by extreme valuation metrics alongside signs of trend exhaustion.
The below post is based on:
Chibane, Messaoud and Samuel Ouzan (2019), “Value Bubbles”.
The below text consists of quotes from the paper, but headings, emphasis, and cursive text have been added for context and ease of understanding.
The post ties in with the SRSV summary on fundamental value.
Value bubbles and momentum crashes
Value investment means buying (or overweighing) securities that rank high on conventional valuation scales, such as book-price ratios or forward earnings yields. Momentum trading means ‘going long’ securities that have outperformed in the recent past.
“Value profits should mirror momentum profits, and vary over time…From 1926 to 2018, following negative market return, the average so-called value premium is about three time its unconditional counterpart, whereas it appears to vanish following positive market return… Risk-adjusted value strategies are not profitable following periods of market gains.”
“We provide evidence of episodes of extreme value profits contemporaneous with market rebound in time of market distress. We call these episodes ‘value bubbles’. We provide evidence that ‘value bubbles’ mirror ‘momentum crashes’…Our findings are consistent with several extended classical overreaction theories.”
“[Academic research] establishes a link between value and momentum premia…[that is] significant co-movements in value and momentum strategies across diverse asset classes and also a negative correlation between value and momentum within and across asset classes…The negative correlation between value and momentum returns is predicted, by several pioneering behavioral theories.”
What governs the dynamics of value stocks in theory
“[It is useful to] disentangle the market risk into permanent cash-flow shocks (bad beta) and temporary shocks to market discount rates (good beta), where investors should care more about the former than the latter and demand a higher price to bear bad beta risk…Value stocks have relatively high bad betas with [respect to] market cash flow shocks…Value stocks tend to co-vary negatively with the expected market risk premium [meaning that they take a hit in market price when the risk premium increases].”
“For ‘behaviorists’…the main argument for the value premium, is that…value investors bet against ‘naive’ investors who may either overreact to news, extrapolate past earnings growth too far in the future, or assume trends in stock prices…Naive investors will tend to follow positive feedback (momentum) strategies, buying when prices rise and selling when prices fall augmenting the profit of value strategies… Value stock returns should therefore increase [as momentum weakens or reverses]. [For example], according to behavioral theory, a decrease in momentum trades dampens the selling pressure from momentum traders on underpriced stocks who have performed badly in the past…[Similarly] an increase in momentum trades strengthens the buying pressure from momentum traders on overpriced stocks who have performed well in the past [reducing] the profit of values strategists.”
“Continuous overreaction leads to positive auto-correlation and subsequent repeated public information arrival to reversal. Since confirming public news furthers the initial overreaction and pushes prices further away from their intrinsic value, the intensity of investors confidence also increases the time taken for mispriced stocks to revert back toward fundamentals…Aggregate overconfidence should be greater following market gain…We should be able to forecast periods when mispriced stocks return relatively quickly to their fundamentals, increasing, ceteris paribus, in these periods the return of value strategies.”
Conditioning value strategies on the state of the market
“Profits of value strategies depend heavily on the state of the market.”
“Following persistent market declines, it is very likely that underpriced stocks have experienced a long series of negative earning shocks. Therefore, past overreaction should forecast superior return for these underpriced stocks…Moreover…conservatism increases following persistent market decline, increasing ceteris paribus the regime transition probabilities, dampening the further possible destabilization of underpriced stocks.”
“We examine therefore the dynamic of value strategies…[and] test whether conditioning on the state of the market impacts substantially their profitability… The data for this [empirical analysis] includes all ordinary common shares on NYSE, AMEX, and NASDAQ. Our sample commences on July 1926 and extends through June 2018. We use value-weighted decile portfolios returns sorted on book-price and dividend-price ratios.”
“We define two states (1)“UP” is when the lagged two-year market return is nonnegative, and (2)“DOWN” is when the lagged two-year market return is negative. We find that the most significant part of value profit is realized following DOWN periods…Value profits are statistically greater following DOWN markets…Then we identify episodes when payoffs of value strategies are extreme [and] find that they are contemporaneous with momentum strategy crashes. “
“[The figure below] highlights…the dollar value of investment following bear markets…We define bear market when the market decreases by at least 20%, without a 20% rally, going back to 1927…Following a sufficiently long-lasting bear market, value strategies…provide a particularly high return…Value stocks outperform the market in all sub-samples.”
“The so-called ‘mirror effect’ [refers to] short periods of extreme losses in the momentum strategy and extreme gains in the value strategy. Market rebounds following panic states are attributed mainly to impressive value gains. Put it differently, the ‘value bubbles’ anomaly reported in this paper appears to be an important driver of the market rebound following market crashes.”