My Library: January 2011

Tuesday, January 25, 2011

High Frequency Trading: natural evolution or another scam?

Published January 25, 2011

By R SIVANITHY

EVOLUTION is supposed to be a natural developmental process which after completion results in the end-product being more advanced, sophisticated and somehow 'better' than what existed before.

In the financial markets, high-frequency trading (HFT) is seen as being part of natural evolution towards greater sophistication. The argument here is essentially 'faster is better' because everyone knows that time is money and speed is of the essence.

This is the preferred portrayal by investment banks, brokers and also exchanges everywhere. The Singapore Exchange (SGX), for example, announced last week it is launching the world's fastest trading engine in August as it looks to compete with other exchanges which have been busy upgrading their trading software.

Most people, however, also acknowledge that some trade-offs are inevitable in the quest for this sort of sophistication. The main one is that retail investors don't have large and powerful super-computers at their disposal and could therefore be disadvantaged. To help them, some form of regulation is needed. So it is that much of the regulatory discussion on HFT revolves around circuit breakers that would halt trading in the event of a high-speed crash.

Retail investors are also offered an alternative. For those not prepared to pit their wits against big computers which can react in a millionth of a second, they are offered passive investment in exchange traded funds (ETFs) or unit trusts, both of which are abundantly available and ever-hungry for retail business.

Our worry is that not enough thought or publicity has gone into just how radically HFT has changed the investment game. Few people realise it but unless fund and ETF managers also have super-sophisticated computers to compete on the same terms as HFT traders, they too could well be on the losing end of most trades.

Zero-sum game

According to a recent news report, traders at JPMorgan and Bank of America last year made a profit every single day for two entire quarters. How is this possible? If you accept that trading is a zero-sum game, then who lost? Most likely it was retail investors and probably a bunch of institutions which don't possess adequate computing power. Is buying unit trusts or passive investing really the ideal solution for the small investor?

To put it bluntly, circuit breakers don't offer much protection in a playing field tilted heavily in favour of HFT traders. And relying on listed, commercially-driven exchanges like SGX to formulate useful policies that would level the playing field would be a mistake because exchanges and regulators themselves are not sure what's best.

Consider that ever since the trend towards demutualisation started a decade or so ago, exchanges - like all listed companies - have been driven by a need to maximise shareholder wealth. This used to take mainly two forms: getting companies to list (and thereby collecting listing fees), and by encouraging active trading (and thereby collecting clearing fees, stamp duties, etc). Both activities were not necessarily mutually exclusive since companies that listed would be happy to see their shares actively traded.

Enter HFT, which has only one goal in mind: to make as much money as possible by exploiting weaknesses and/detecting trends before anyone else. (In the US, the average speed of a trade on NYSE Euronext is now reported to be 0.7 of a second or less, compared to 10.1 seconds in 2005 - when it was still the NYSE - while the average trade size has dropped from 724 shares to 268 shares.)

HFT traders have been described as 'valuation agnostic' - that is, they are indifferent to fundamentals or valuations and are concerned only with exploiting money flows and weaknesses.

Short-term trading

There are many strategies employed - for example, some use 'pattern recognition' (which may well be a euphemism for front-running) to detect changes in institutional flows and to buy or sell ahead of those flows while others earn rebates from exchanges for trading in high volumes; so the more liquidity they generate, the better.

However, all seek to square off their positions by the end of each session, so there is as little overnight exposure as possible.

In a nutshell, with HFT, stock markets are now predominantly populated by very short-term traders who can identify trends faster than most others and can react before the blink of an eye.

Profit-driven exchanges are keen to encourage this sort of activity because of the volume it creates and the money it adds to their coffers. But, in time, this can only lead to a near-total marginalisation of the retail investor.

More HFT may also lead to fewer good-quality companies listing since the activity does not concern itself with fundamentals - only with trends, liquidity and momentum. If so, a possible scenario is that, over time, mainly speculative-grade firms will be taken public or remain listed - which would then aggravate the short-termism already in place.

How might profit-driven exchanges like SGX then reconcile the need to ramp up trading volume by having more volume with their other role of encouraging more listings?

At this time, it looks very much that HFT benefits mainly the few - namely the large investment banks which have access to extremely powerful computers.

When you think about it, these are the same people who brought about the 2008 sub-prime mortgage crisis when they scammed the world with their new and 'improved' structured products, all of which eventually collapsed.

So does HFT really equate with evolution and its associated 'new and improved' connotations? Or is it simply another scam peddled by the big houses which have to find ways to keep their profits up?

Or, more insidiously, does one naturally lead to the other?

Monday, January 10, 2011

In search of good long candidates

The criteria would be a low price-to-sales ratio, a high Piotroski F-score and capital discipline.

Mon, Jan 10, 2011
The Business Times

By Teh Hooi Ling
Senior Correspondent

BACK in June 2009, I spun a report off an idea by one of my favourite analysts, James Montier, who was then still with Societe Generale.

Mr Montier published in May 2008 a report titled 'Joining the Dark Side: Pirates, Spies and Short Sellers'. He screened stocks for three characteristics: a high price-to-sales ratio; deteriorating fundamentals (as measured by a low Piotroski F-score - I'll elaborate later); and poor capital discipline (as measured by high total asset growth).

'Each of these characteristics is a telling sign - but when combined, they become even more potent,' Mr Montier wrote.

According to him, over the period 1985-2007, a portfolio of such European stocks rebalanced annually would have declined more than 6 per cent per annum in absolute terms; in comparison, the general market was rising 13 per cent per annum. The basket of stocks suffered absolute negative returns in 10 out of 22 years. It underperformed the index in 18 of the 22 years. Similar findings hold for the US, Mr Montier said.

However, in the few years leading up to 2008, this strategy totally backfired. The basket of stocks that one was supposed to short, based on the above criteria, actually outperformed the general market. 'This attests to the extreme nature of the dash to trash that we had witnessed,' he said then.

In May 2008, he ran the screen and saw a record number of stocks passing the criteria for the short basket. Generally, such a screen would yield 20 stocks in European markets. But when he looked, the number of stocks totalled 100. In the US, the typical average number would be 30 stocks; in 2008, however, it was 174. 'The opportunities are on the short, not the long side, currently,' concluded Mr Montier. 'Perhaps it is time to join the dark side.'

He was spot on, and the ensuing global financial crisis totally validated his view.

In my June 2009 piece, I noted that prices had recovered somewhat, but were still some 30 per cent below the levels in May 2008. 'And the economy is supposed to be on the mend, perhaps we could try to do the reverse of what he has done to see what names we come up with as potentially good long or buying candidates,' I wrote.

Valuation, fundamentals and capital discipline

Here are Mr Montier's criteria. First is valuation. From a short-seller's perspective, the most useful valuation measure is price-to-sales. According to Mr Montier, high price-to-sales stocks allow investors to hone in on counters that have lost touch with reality.

For example, in 2000 Sun Microsystems was selling at US$64 at 10 times its revenue. Its then CEO noted in 2002: 'At 10 times revenue, to give you a 10-year payback, I have to pay you 100 per cent of revenue for 10 straight years in dividends. That assumes I can get that by my shareholders. That assumes I have zero cost of goods sold, I have zero expenses, and that I pay no taxes.

'And that assumes that with zero R&D for the next 10 years, I can maintain the current revenue run rate. Now, having done that, would any of you like to buy my stock at US$64?'

For Mr Montier, a company would fall onto his short list if it has price-to-sales ratio of more than one.

The second short criterion is deteriorating fundamentals.

Here, he uses Joseph Piotroski's F-score. There are nine measures that relate to a company's profitability, the health of its balance sheet and its operating efficiency that one looks at to arrive at the F-score. The checklist is: positive net earnings; positive cash flow from operations; increasing return on assets (ROA); higher operating cash flow relative to net income; decreasing long-term debt as a proportion of total assets; increasing current ratio; stable or decreasing number of shares outstanding; increasing asset turnover; and increasing gross margins.

Every yes answer to the items on the check list gets a score of one, and no a zero. So companies with improving fundamentals will have higher F-scores (seven to nine), and those with deteriorating fundamentals will have low F-scores (below three).

Piotroski's strategy calls for buying a company with the requisite low price-to-book (P/B) ratio and an F-score of eight or nine. Piotroski's research shows that low P/B stocks with high rankings are less likely to go bankrupt or to fall drastically in price than those with low rankings. This provides a greater safety margin.

Mr Montier's final criterion is capital discipline. Companies that have grown their total assets by more than a double-digit percentage will pass through his short screen.

So for my 'buy' or 'long' list, I set it such that the companies should have a price-to-sales ratio of below 1.2 times, Piotroski F-score of more than seven and total asset growth of less than 20 per cent from a year ago.

At that time, the lone stock the screening process spat out was Singapore Petroleum Corp which was then being acquired by PetroChina.

I then relaxed a few of the criteria and the stocks which Bloomberg came up with and which I listed in my article then were: SembCorp Marine, Keppel Corp, Cosco, Venture, SembCorp Industries, Yangzijiang, Noble Group and Thai Beverage.

Now one and a half years down the road, how would one's portfolio had done if he or she had bought all the stocks listed above?

The performance

As it turned out, four out of the eight stocks on the list were in the offshore and marine sector. Yangzijiang emerged as the best performer - it returned 157.6 per cent between June 22, 2009 and yesterday. Dividends received for all stocks were assumed to have been reinvested at 0.37 per cent a year. Noble Group managed 103.7 per cent, and SembCorp Marine 97.5 per cent. (See table)

The worst performer was Thai Beverage which returned 45.7 per cent - that's just a shade lower than the Straits Times Index ETF's 47.9 per cent return.

On average, the basket of eight stocks returned 86 per cent and outperformed the STI ETF by a whopping 38 percentage points!

So the screening process proved pretty effective. Yesterday, I did the screening again. The lone stock which met all the above criteria, ie full F-scores of nine, trading below 1.2 times sales and below book, and registered less than 20 per cent increase in total assets over the past year, is Time Watch.

And if I relaxed some of the criteria, the list expanded to include Spindex, GP Industries, PCA Tech, Chosen, Pertama, Beyonics and CFM Holdings. This time round, the list is made up of mostly small cap stocks. It seems the market recovery has lifted most if not all bigger cap stocks out of the 'cheap' zone.

I also decided to run the screen on the other regional markets. For markets like Taiwan, China, Indonesia and Vietnam, there were no stocks left despite having seven more criteria to go through. This could suggest that these markets are more exuberant than others.

In Japan, the only stock left standing was Taiyo Elec Co, a company which develops, manufactures and sells pinball machines used in pachinko parlours.

In Korea, if I removed the capital discipline screen, Pungkang Co Ltd is the only stock which showed up. The company manufactures and sells automotive parts like weld nuts, lock nuts, flange nuts, wheel nuts, plate nuts, conical nuts and hex nuts to automobile manufacturers such as Hyundai, Kia and Daewoo.

Readers who may be tempted to act on the above names are, of course, advised to carefully evaluate the fundamentals of the stocks yourselves, particularly given that almost all on the list are small companies stocks which inherently face a tougher operating environment out there.

The writer is a CFA charterholder

This article was first published in The Business Times.