earlier)ranasimilarregressionwith 106 privateplacements
between 1980 and 1987 and also found larger private
placementdiscounts atmoredistressed,riskier,andsmaller
firms.
Theseregressionsarealittlemoredifficulttoadaptforuse
with private company valuations since they arecomposite
regressionsthatincluderegisteredprivateplacements(where
there is no illiquidity). However, the results reinforce the
Silber regression findings that troubled or distressed firms
should have larger illiquidity discounts than healthy firms.
Therearelegitimatecriticismsthatcanbe mountedagainst
theregression approach. Thefirst is thatthe R-squared of
theseregressionsismoderate(30to 40 percent)andthatthe
estimates will have large standard errors associated with
them. The second is that the regression coefficients are
unstable and are likely to change over time. While both
criticismsarevalid,theyreallycanbemountedagainstany
cross-sectional regression and cannot be used to justify a
constant discountfor allfirms. After all, these regressions
clearly rejectthe hypothesis that the discountis the same
across all firms.
Synthetic Bid-Ask Spread
Thebiggest limitationof using studies based on restricted
stockorprivateplacementsisthatthesamplesaresmall.We
wouldbeabletomakefarmorepreciseestimatesifwecould
obtainalargesampleoffirmswithilliquiditydiscounts.We
wouldarguethatsuchasampleexists,ifweconsiderthefact
thatanassetthatispubliclytradedisnotcompletelyliquid.In
fact,liquidityvarieswidelyacrosspubliclytradedstock. A