• Sun
  • Aug 24, 2014
  • Updated: 3:54am
PUBLISHED : Tuesday, 22 October, 2013, 3:27am
UPDATED : Tuesday, 22 October, 2013, 11:31am

Hong Kong's stamp duties are woefully inefficient ways to tax

The argument that revenue raised from taxing share transactions is a pay cheque for Hongkongers for keeping the rule of law is short-sighted


As the writer of the South China Morning Post’s Monitor column, Tom Holland attempts each day to make sense of the latest developments in business, finance and economic affairs in Hong Kong and mainland China.

On Sunday, my esteemed colleague Jake van der Kamp duffed me up in print.

I'd provoked his ire by arguing that the Hong Kong government should scrap its stamp duty on share trading.

Jake disagreed.

He accepted my points that the government's 0.2 per cent per round trip tax on trading reduces liquidity, eats into savings and bumps up the cost of capital.

But on balance, Jakes believes the price is worth paying.

The mainland tweaked its stamp duties … with a marked impact on trading volumes

"This particular stamp duty is Hong Kong's reward for maintaining rule of law," he declared in his Sunday Morning Post column.

Jake pointed out that when investors buy shares in Hong Kong, they can be confident their ownership rights will be respected. And if anything does go wrong, they have recourse to a judicial system that has a reputation for being both honest and fair.

In a nutshell, he argued that Hong Kong's system works, and that the HK$20 billion in stamp duties investors handed over to the government last year are a fair price to pay for its efficiency and reliability.


Sorry Jake, I don't buy that line. I agree that Hong Kong's system costs money, and accept that the government has to raise revenue to fund it. The trouble is that stamp duties are a woefully inefficient way to raise revenue.

There are plenty of examples to illustrate this.

In 1986, Sweden doubled its stamp duty on share trading. Over the next few years volumes on the Stockholm stock exchange plunged as more than half of the total trading in Swedish equities migrated offshore to London. As a result, the Swedish government was left worse off.

The effect on the fixed-income market was even more pronounced, as bond trading volumes plunged by some 85 per cent.

China, too, has found that stamp duties are a lousy way to raise revenue.

Since 1997, the mainland has tweaked its stamp duties on stock transactions seven times with a marked impact on trading volumes.

For example, in 2007, when the authorities raised the duty per round trip from 0.2 per cent to 0.6 per cent, onshore trading volumes in the shares of companies with both mainland and Hong Kong listings collapsed relative to their offshore counterparts.

According to one study, the mainland experience indicates that a 0.22 percentage point increase in stamp duty tends to trigger a 28 per cent fall in trading volumes.

But you don't even need to go as far as the mainland to see that stamp duties are an ineffective way to raise tax revenues.

In February, the Hong Kong government doubled the stamp duty it charges on the sale and purchase of a typical apartment in the city from 3 per cent to 6 per cent. We don't yet have the official data to show what effect the increase had on stamp duty revenues.

But we can make a pretty fair guess.

As the first chart shows, since the stamp duty rise, the volume of residential transactions has slumped from about 8,000 a month to 4,000.

With prices remaining stable, that means the value of taxable transactions has fallen from about HK$40 billion a month to HK$20 billion. So the doubling of stamp duties has halved transaction values, which means at best the government's move won't have raised any extra revenue at all.

And if you factor in the higher rates on more expensive properties, and punitive duties on quick sales and on buyers who are not permanent residents, then it is likely overall revenues from stamp duties on residential property transactions have actually fallen as a result of the government's rate increases.

So, sorry, Jake, if it's a matter of raising revenue to pay for the rule of law in Hong Kong, stamp duties are a lousy way to go about it.



For unlimited access to:

SCMP.com SCMP Tablet Edition SCMP Mobile Edition 10-year news archive



This article is now closed to comments

The columnists argue from different value perspectives
each argument seems valid from its premise
both make enjoyable reads
and help justify continuation
of my not-entirely-willing support for scmp
Holland, to shore up his views, have inserted property transactions (tax) into the equation. But it is a totally different kettle of fish!!!
Does Holland suggest other industries subsidise this lucrative sector? Also, Giwaffe's comments is spot on - the share transactions are pricey because of the spreads and high commision/fees charged by the brokerages.
I think Holland needs to acquaint himself more with the whole issue (and finance industry) before attemping to put forth wishy washy "arguments." Jake was correct to correct him!
Tom, your argument seems to be that increasing stamp duties will not lead to an increase of government revenues. Quite the opposite, it will lead to a decrease of transaction volumes and corresponding government revenues as highlighted in your article. This all seems to make sense.

However, this does not automatically validate the opposite argument, which is decreasing stamp duties will lead to an increase in government revenues. The Laffer curve comes to mind here. At the presently low rate of 0.1% of the transaction value, the stamp duty is an infinitesimally small amount of tax. It strains credulity that reducing or even abolishing this woefully negligible amount of stamp duty could have any material effect on the equity markets. In fact, I might even argue that the stamp duty rate is below its optimal and may even be raised to increase government revenues.

Instead, I would argue that the exorbitant costs of equity trading (as compared to other financial markets) from relatively higher commissions, brokerage fees, exchange fees, and other levies in Hong Kong are the culprit. For example, online trading of 1000 shares of HSBC would cost approximately HKD 200 to 250 at most brokerages. For comparison, the same trade would cost USD 7 to 10 in the US markets, or HKD 55 to 80.
I would argue that the high minimum bid ask spreads increase the cost of trading and discourage transactions. How does it make sense to have a minimum spread of HKD 0.05 for a stock priced at HKD 5.00? In the US markets, the spreads can be USD 0.01 for stocks priced over USD 50! Minimum spreads exist mostly to safeguard profits for market makers (sometimes called liquidity providers).
Yes, taxing anything increases its price.

This will lead to a fall in demand (but not an increase in supply; the government captures some of the surplus), and hence a decrease in quantity/liquidity. Thanks Einstein.

Taking the above argument to its conclusion, taxing anything is inefficient: labour, property, all kinds of transactions, consumer goods, company profits etc.

Mr Holland's argument is hollow. Unless he wants to abolish government completely, we need to tax something. And whatever we tax, will become less 'liquid.' A stamp duty on stock transactions of 0.1% is an entirely reasonable part of a fiscal regime.


PS. Sweden 1986? Please. They doubled their stamp duty to 1.5% one-way, so 3% return. Yes, compared to London's 0.5% one-way at the time, that made a full 2% difference. That is quite a push factor indeed if you are a big boy and trade millions. Hong Kong's 0.2% return charge is peanuts in comparison.

For Hong Kong property, we are talking about stamp duty rates well into double digits in most case now. How is this possibly comparable to the 0.1%? Same thing for the mainland example of stamp duty to 0.6% return, that might (note: "one study" says it "indicates") have decreased volume. And all of that is at best an argument for not increasing the stamp duty beyond some optimal point where tax revenue is maximised from a volume*price perspective.

None of it pleads in favour of abolishing it, when such revenue would fall to zero.
@jve: Great analysis!
[0.22 percentage point increase in stamp duty tends to trigger a 28 per cent fall in trading volumes.]

So when the mainland increased stamp duty from 0.2% to 0.6%, this lead to a drop in trading volume of what? About 50%?

They increased the stamp duty by 300% and volume went down 50%? Then they still came out on top from a tax revenue perspective. That tax revenue increased by a massive 50%. So what is your point Mr Holland?
Good supports, TH. You win this round.
Tom has totally missed the boat. He, too, is not infallible. This is one such case.
John Adams
I admire Jake's independent thinking on issues, but he is not infallible.
I think you out-gunned him on this one Mr Holland




SCMP.com Account