issue shortly—rational investors will demand compensation for it. In other
words, they will require that the fund’s shares trade at a discount to NAV.
This also explains why new closed-end funds are often sold at a pre-
mium. Entrepreneurs will choose to create closed-end funds at times of in-
vestor exuberance, when they know that they can sell fund shares for more
than they are worth. On the other hand, when a closed-end fund is liqui-
dated, rational investors no longer have to worry about changes in noise
trader sentiment because they know that at liquidation, the fund price will
equal NAV. They therefore no longer demand compensation for this risk,
and the fund price rises towards NAV.
An immediate prediction of the LST view is that prices of closed-end
funds should comove strongly, even if the cash-flow fundamentals of the as-
sets held by the funds do not: if noise traders become irrationally pes-
simistic, they will sell closed-end funds across the board, depressing their
prices regardless of cash-flow news. LST confirm in the data that closed-
end fund discounts are highly correlated.
The LST story depends on noise trader risk being systematic. There is
good reason to think that it is. If the noise traders who hold closed-end
funds also hold other assets, then negative changes in sentiment, say, will
drive down the prices of closed-end funds andof their other holdings, mak-
ing the noise trader risk systematic. To check this, LST compute the correla-
tion of closed-end fund discounts with another group of assets primarily
owned by individuals, small stocks. Consistent with the noise trader risk
being systematic, they find a significant positive correlation.
6.2. ComovementThe LST model illustrates that behavioral models can make interesting pre-
dictions not only about the averagelevel of returns, but also about patterns
of comovement. In particular, it explains why the prices of closed-end funds
comove so strongly, and also why closed-end funds as a class comove with
small stocks. This raises the hope that behavioral models might be able to
explain other puzzling instances of comovement as well.
Before studying this in more detail, it is worth setting out the traditional
view of return comovement. This view, derived from economies without
frictions and with rational investors, holds that comovement in prices re-
flects comovement in fundamental values. Since, in a frictionless economy
with rational investors, price equals fundamental value—an asset’s ration-
ally forecasted cash flows discounted at a rate appropriate for their risk—
any comovement in prices must be due to comovement in fundamentals.
There is little doubt that many instances of return comovement can be ex-
plained by fundamentals: stocks in the automotive industry move together
primarily because their earnings are correlated.
48 BARBERIS AND THALER