I have generally been on the side of the Free Market Barons against the Robin Hood tax in the blogosphere discussions so far.  This does not mean I think it is simple.   The problem for me is that the argument is largely misspecified. As James Macintosh demonstrates, many of the charities that are behind the tax don’t seem to understand what they are on about.

Start with the advert. It seems to imply: “this is infinitessimally small, this tax.  0.05%!   Compare that to 17.5% on VAT!.  So surely uncomfortable looking banker Bill Nighy, you could not complain about this itsy bitsy thing?” But it is not itsybitsy – it is huge. To understand why, look at this graph plucked from random about the £-$ rate last Friday:

For a typical retail punter, in this glorious age, you can deal on this price as a speculator with a spread of 5 pips.  (for an example, see the prices moving here: disclosure, I worked there for a long time).  Although for a bank the spread might be tighter, this may depend on the size – shifting $100m and the spread might be 5 pips again.

Most speculators open or close their positions well within the time period shown on this graph.  And observe: the price moves from 15600 to 15640.   If you paid a 0.05% charge round-turn, or 8 points in this case, then you are wiping away most of the profit opportunities.  You would have to be fairly convinced that you have a 10 point advantage over the market in order to win.  As it turns out, 10 points is Don Bradman territory – in current markets .  To consistently out-guess markets to that degree would normally involve an edge that bordered on criminal.

So the point about the 0.05% is not to divide through by the entire value of the currency but by the reasonable range of its variation. In this case, more like 1% (In fact, most speculators would enter a trade with a stop-loss much closer than this.).  So, for most short term speculators, it is the equivalent of taking a project of some kind or other, and charging 5% of the revenues just to start up.

To conclude this first part: this is not a tax on something that is in any way comparable to, say, a shopkeeper taking a product, marking it up and selling it.  It is a tax on speculation, but at these levels, it is something even more: it is a virtual outright ban on any speculation that is not looking for very large average profits.   Consider: a typical trade for a retail punter might be to buy £500,000 £/$.  0.05% of this is about £250.  The typical brokerage fee for a futures trade might be $6.  If that £500k is 10 contracts, the itsybitsy tax is in fact about 6 times a broker’s charge. No industry survives that – imagine, say, that there was a £50 tax on cinema tickets.

So what?  Why are we sitting around discussing how to make life easier for people betting on a number? How on earth can the ease with which one irrelevant reptile beats another irrelevant reptile, at their irrelevant-but-possibly-dangerous game of poker, matter in any way whatsoever?  Several answers here:

  • Even if it did not matter, if the tax were fully enforced as imagined, the people proposing it can clearly have no idea what it would do to volumes – and therefore revenues.  To say “cratering” is an understatement; for FX alone you are clearly trying to remove a high multiple of the profits  already available**.  It is like trying to guess what the yield from a £1-per-email tax would be.*  So: no revenues anywhere near those proposed
  • What would remain? Well, FX (I’m sticking with FX for simplicity) would remain in some form.  The vast majority of punters interested in sub 50pip movements would be gone.   But there would still be profits to make.  In fact, there would be VAST profits to make – because the market would suddenly be really inefficient.  Because no-one would be trying to pick off small anomalies, small under-reactions to news, the market would be illiquid, jumpy, unreliable, gappy.   (A bit like the CDO market, in August 2007).    Instead of it being a market with very high volumes and low per-trade profits, there would be terribly low volumes and large per-trade profits.
  • In many ways, this is far more fun for finance.  You can justify all sorts of other charges, when the perils of dealing in the market are so high.  Of course, you have a price to pay the government, but you have a large gappy market within which all sorts of other practises can take place.
  • But the market would still be illiquid.  And this is where the real cost lies.   Currently, there are hordes of small speculative positions, and then giant ‘end users’ of finance, who need to convert risks from one form to another.   It is those end users that would really take a hit.   Instead of facing a 0.2% spread for doing a gigantic FX transaction, the absence of millions of small other-speculators on the other side might blow this out – to 0.5%, to 0.9%, more.    Who knows?  But this translates into a giant cost of capital charge on everyone else.

The real point I wanted to make from this rant is: don’t try to address the potential costs of doing this by using the wrong model.  The ‘wrong model’ pictures banks doing something – a metaphorical equivalent of digging gold out of the ground – and currently passing it on to a customer.  This proposal, under the wrong model, imagines an itsy bitsy extra charge of 0.05% on this. How could that hurt?  The ‘wrong model’ also goes onto to imagine that the banks take all this charge, meekly cutting their profits.  Hence ensues a largely irrelevant argument on LibCon between Sunny and otheres about whether they would or not.   This is missing the point.

Instead, what this proposes is to raise a much smaller amount than suggested, in return for which an entire field of speculative activity would be laid waste.   Masses of markets reliant on the underlying spot market would be wiped out – options need a liquid spot market, for example.   As a result, the real firms and businesses, from insurance to industry to government finance, would face a massively higher cost. Their costs would rise, in all sorts of difficult to understand ways.  They would hold back from some opportunities that they would have otherwise exploited more fully.  They would raise some prices higher than they might have.  Some new entrants will be put off, so industries would be less competitive.

The consequences would be very hard to predict, against an equally-hard to predict benefit.  I don’t know where the trade-off lies.  Like James again, I think finance needs controlling, but in other ways; this is the equivalent of asking it to work with inferior technology, or lobotomised staff. It is, in short,  a deliberately induced supply shock on the whole economy, like a sudden loss of technology.  Not a decision to take lightly, or with the wrong mental model of what you’re doing …..


I’ve broken my word limit.   NO time to make it shorter.  But there are good come-backs like – spreads were far larger in the 1980s and we still grew then – so what is the problem?

*If some bright spark thinks “a lot”, then ask yourself why you still don’t think it is a good idea

**My best guess is that if banks make $20bn annually from FX, then that is about 0.005%, or a tenth of the proposed charge.


20 thoughts on “The Robin Hood Tax

  1. I don’t think it is a good idea, and this is a good post, but I wonder about this point (which is the crux of your argument):

    In fact, there would be VAST profits to make – because the market would suddenly be really inefficient.

    As far as I can see the IG Index spread you link to is 0.038%, so its about 3/5th of the proposed tax. Now that’s for retail customers and I assume wholesale get smaller spreads. But what what were they ten years ago or even 20 years ago? If they were substantially larger than that then did have a huge negative effect on other users of forex markets and were the markets really inefficient?

    1. this is a point I am struggling with, to be honest, and it is one I am unsure about giving a verdict on. My feeling is that rolling back to an earlier level of inefficiency is almost certainly bad news, even if people at the time coped well. Consider: in 1980, there was no electronic communication. we got by fine, in terms of being able to grow. But would wiping away electronic communication now be a good idea? Clearly not. Most of today’s industrial structure needs, to some extent, finance to have a particular cost. Think how globalised trade is now. How much would be disrupted if it cost 10 X as much to arrange finance?

      I am not sure of the answer and your question is one I will have to sleep on….

  2. I have to admit, I’m out of my depth with this stuff – I have no real knowledge of the world of people making large bets for tiny (in percentage terms) returns, and I don’t even really understand how anybody makes money, regularly, from FX speculating in the fashion you describe. I don’t know what the composition of FX trading is – I mean, if you broke it down into who’s doing what and why, I don’t knwo what that’d look like.

    What about people who take bets on (or against) currencies where they are expecting large movements and are prepared to wait long periods to see them? How about carry trades, investments in overseas government debts? Isn’t there enough of that going on to provide liquidity, in the sense of still being able to buy or sell whatever quantity of whatever currency at the going price, without (in most cases) your trade being large enough to move the price itself? Aren’t importers and exporters still going to be out there every day, buying and selling currency? And bond investors? Aren’t firms going to be selling (or buying) currencies forward with intermediaries who then going and buy (or sell) the underlying currency?

    I guess what I’m saying is that I can understand how – just in the context of the FX market – a 0.05% tax is going to raise the cost for everybody of doing business (by widening the spread) and I can understand how it will wipe out certain kinds of activity (those for which 0.05% is a big number) but I can’t really follow the argument that the tax will erase speculation altogether, nor that it will create a seriously illiquid and inefficient market. After all, Tobin described it as throwing sand in the wheels, not a spanner… did he really not understand the implications of what he proposed?

    1. On how or whether they make money, I am happily agnostic. I used to make money more as a market maker than a trader. But all sorts of weird chains and linkages can take place. Long term I think some people make more money out of others because the former have the choice and means to be more discriminating.

      For example: big Bank offers a put option on the £/$ to a £ exporter, to sell £$ at 1.50 in a year. This enables the £ exporter to plan forward with confidence into the US market. The £ exporter pays 4c on the deal i.e for every £1m he is paying, I think, $400. The Bank pocket this $400 but now has a problem; if the £ falls to 1.50, he is on the hook for £1m sterling falling in value. So he delta hedges, selling, say, £100k at 1.5600. Given his knowledge of liquidity and likely volatility, he will make money on the option.

      That sale at 1.56 is absorbed by a market that probably thinks the fair price of the £ at that second is 1.5601, so making $10 for the speculator on the other side who may be buying for entirely different reasons. If he were not there, because of a tax, the spread for the Bank’s actual size (let us multiple everything up so the deal is in fact for £1bn) might be 1:5570 at 1.5630. Given this fact, the Bank might have sold the option for 5c or 6c per £million, costing the exporter much more -for the same risk mitigation.

      I don’t think it would erase speculation altogether, but that some outright costs like the option above would become way worse. Tobin seemed to have a far less benign view of speculation – as driving currencies away from fair value. I just don’t see a million people taking 40-pip views as doing that.

      But I think we aagree about the fundamental point of my overlong post: it is not about tax incidence in a straightforward way, but about whether a massively trade constricting levy can raise enough money to justify the enormous costs this throws on users.

      1. I just want to pedantically make sure I really understand what you’re saying, because this is the first time I have encountered this … if I follow you, the idea is that the full chain of transactions involved when a firm does something useful to the real economy, like hedging its currency exposure, is such that a small charge in one place would be amplified into a large change in the end cost of this service. This is for FX – potentially there may be analogous mechanisms in other areas, such as hedging interest rate risk or whatever. Of course these transactions would still take place – only now just when the benefit of risk mitigation outweighs the higher cost of doing so. So the end result would both be higher costs to firms (and hence consumers) and greater risk exposure (all those instances where the higher cost means risk is borne rather than hedged).

        It’s not easy getting a handle on how important the above would be … is that really an “enormous cost” in the scheme of things?

      2. You have characterised my position fairly, except I would take dispute with ‘one small change in one place’. It is a huge change, equivalent in scale to putting a charge on emails, I think, because of the volumes involved. I am struggling to find a really decent analogy here.

        Very hard to get a handle on the consequences. I believe that there have been many macro studies linking higher cost of capital/liquidity to lower growth. But you’d have to ask someone better plugged into the academic work, like Chris Dillow. As an economic historian, I tend to revert to France in 1800 (notaries) and Britain in 1800 (banks and markets). not sure it would convince Bill Nighy

  3. I do recognise, though, that the argument you are making here is fundamentally different to the “tax incidence” argument I was having with Sunny, which I regret not only because I became needlessly unpleasant, but also because it is beside the point: because it applies to all taxes, it doesn’t say anything useful about the wisdom of any particular tax. It just bugs me when people deny it. Here you’re not just saying that the tax will be – partially – borne by businesses and consumers, you’re saying it’ll do harm to the functioning of the real economy. Which is why I want to understand your argument a bit better.

  4. I’ve noticed a major problem with the above analysis (and for all I know the original paper’s estimate of revenue raised). The proposed tax in the Richie paper and being promoted by the Tuc is 0.005%, not 0.05%, ie half a basis point not 5 basis points. I think this would make most of the concerns about reducing liquidity much less appropriate – this would be only about 1/6th of the IG index spread.

    But the Nighy advert and Robin Hood campaign is 0.05%.

    So we’re.they’re talking about different things.

  5. This piece is based on a currency transaction tax of 0.05 per cent.

    While it won’t probably change your mind, the suggested rate on currency transaction is 0.005 per cent.

    The Robin Hood campaign’s campaign is for an average of 0.05 per cent, but that is based on a spread of rates between 0.5 per cent on shares and 0.005 per cent on currency and other low-margin high volume trades such as derivatives.

    You could argue that the ads oversimplify this, but that is the nature of ads – and a different argument.

    The TUC, Christian Aid IMF submission which is much more detailed than any asd can be is now available at


    1. Nigel, Matthew, many thanks for the clarification.

      It no doubt changes matters. Whether decisively or not is another matter. The question should still be specified in terms of whether $x of raised cash is a price worth paying for $y of reduced liquidity – not ‘do you feel like takign this much from these Banker Barons in return for them having a really minor inconvenience’ – but I understand the imperatives of modern campaigning….

      1. One other point concerning my suggestion that it would take us back to the position 10 or 20 years ago in terms of liquidity (perhaps now its 0.005% not 0.05% it would only be a few years ago). Isn’t it the case that academic studies found that forex markets were ‘efficient’, in that you couldn’t reliably make money from anomalies, in the 1980s and early 1990s? Or is that completely not the case?

    2. Thanks Nigel

      By the way, am I missing something, or are you missing something: banks do not currently pay stamp duty trading shares, surely? Your own figures suggest that about 11 trillion of shares are traded in the UK each year, and at 0.5% that would raise 50bn a year – but stamp duty – INCLUDING HOUSES – raised just #5bn in the last estimate. So on shares it raises, what, #1bn, 2bn?

      Your report says

      “There appears little doubt that trading is not being prevented by tax charged
      in that case, albeit the size of the non-order book market suggests that derivate deals are high and rising as a

      Most of the trading is not taxed at 0.5%. And by the way the non-order book market does NOT mean derivative deals! There are two ways of trading shares, SETS and SEAQ . .. . I don’t want to be picky, but I am not sure the author is an expert ….

  6. Dear Giles

    You seem to be unaware that “speculators” are all evil, blood-sucking parasites that perform no useful economic activity whatsoever. They only exist because bankers are greedy.

    Please compare with: “real investors”.


    The Arrogant-lef-that-doesn’t-know-anything-about-economics-because-we-already-know-it’s-all-lies-so-we-don’t-need-to-know-about-any-of-it

  7. “By the way, am I missing something, or are you missing something: banks do not currently pay stamp duty trading shares, surely? ”

    They do not as market makers: when they do so in that capacity they are not treated as acquiring the share as such but as acting as agent – just as, for example, a second hand car dealer is not registered as owner of a car on their forecourt

    So what? This does not change the data. The UK data you refer to is consistent with the data on total revenues noted, but of that total data Dean baker estimates $165 bn would be from the USA

    That does not require the UK to raise the revenue you are seeking to extrapolate – far from it

    In other words, we’ve noted and checked this point before publishing – but thanks for providing the chance to offer clarification

    1. Hi Richard

      Thanks for coming back.

      What I tend to question, then, is how much the UK precedent proves about the effect on liquidity elsewhere of this proposal. Are you proposing that only non-market maker activities are to be taxed in future – and do you have a breakdown of global share trading between market makers and traders? Effectively, the way stamp is constructed in the UK, it is only end-users of the shares – i.e. investment funds, normal puntes – that pay it – not the bankers ceaseless trading the shares to and fro, in the way that the TUC et al find so dangerous.

      So either you are taxing the ‘dangerous’ market makers high velocity activity – in which case liquidity will dive bomb – or just the end users – in which case this is not really about bankers. ….



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s