David Brooks writes today in the NYT about economist and co-director of the Center for Economic and Policy Research Dean Baker's proposal to create a "financial transactions tax."  This tax (the writer suggests .25%) would be assessed on any sale or transfer of stocks, bonds or other financial asset.  The columnist does not specify whether the seller or the buyer would pay the tax, but in a Coasian world it does not matter.  (Part of the article seems to suggest it would be both -- by hypothesizing that someone would "pay a quarter-percent fee to purchase the asset in the first place and then another quarter percent to sell it."  I'm not sure if this was intended or not.  If so, then there is a total of a .5% fee every time a security changes hands.)  This fee is projected to raise perhaps $100 billion a year.

Of course, a lot of different new taxes would raise a lot of money -- that's what taxes do -- but we should try to anticipate how a tax (or a deduction) would skew behavior, and try to determine whether that is how behavior should be skewed.  The unintended consequences of varying tax policies may in fact add to economic troubles or just create new ones.  Many have argued that the mortgage tax deduction may have been an accomplice in the housing bubble, so let's be a little careful before we start imposing taxes or subsidizing behavior with deductions.

My first reaction at reading the first part of the article was that if we want the capital markets to be efficient, taxes lead us away from efficiencies.  (Remember one of the (false) assumptions of the efficient capital market hypothesis is that there are no transaction costs.  When taxes are present, studies show that markets move away from efficiency, and this is why we see "January effects" for certain types of investments and "November effects" for others.  Adding a new tax to buying and selling securities seems like it has the tendency to move the market away from efficiency.  At this point in time, we want trading, right?  The tax would have to be fairly nominal if it is not going to skew behavior.  Is .25% nominal enough?  Maybe.

Except that Professor Baker and David Brooks want the tax to skew behavior -- speculative behavior.  You see, it's the really bad people known as speculators who are ruining our markets, who "bring a manic quality to the markets, who treat it like a casino."  Bad speculators trade multiple times a day, so the small fee would add up and possibly make them just calm down and trade slowly, like the rest of us prudent investors.  The "beauty" of the tax is that it is a "progressive" tax "that discourages nonproductive activity."  Hopefully some tax people will jump in here, but I think of progressive taxes as those that affect people more as income increases.  Do we know that day traders or speculators have more income than "buy-and-hold investors"?  We like to say that the "buy-and-hold" people come out on top, so then isn't the tax regressive -- taxing the silly speculators who don't understand investing, just like lotteries are regressive taxes on poor people who are bad at math?  I guess the speculators we really, really hate are the ones that make money.  Bad speculators.

So, the second question is whether speculation really has no utility or even negative utility to the market.  Studies are just appearing telling us what happened when short-selling was banned last Fall, so maybe we'll have the answers soon.  But regardless of whether you are a speculator, prudent investors need speculators for liquidity, and issuers need the presence of liquidity to be able to sell publicly-traded shares in the first place.  If we tax speculators away, then our financial models that assume the ability to sell a particular security at some price may have to be tweaked once the buyers are gone.

Securities, Taxation | Bookmark

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345157d569e2010536ccd9ec970c

Links to weblogs that reference Taxing Speculation: