The first point to emphasise is we will always be Asian specialists. Previously – before setting up OQ – we diversified outside of Asia. This was partly driven by commercial considerations. The team was large, the Asian fund had reached capacity, and we wanted to increase AUM.
Focussing on one region makes sense given the discretionary overlay. It’s important to understand the stock universe. Although we don’t do fundamental stock analysis, we need to understand if there are any issues which could undermine the predictive power of our systematic quant investment process.
Structural, management, corporate action, cross shareholding, regulatory and embargo issues are all relevant.
Although this is a long list of potential issues, we can reduce the workload by focussing on:
Deep value stocks which look “too good to be true”,
Stocks with sell recommendations (although consensus recommendations have little to no predictive power, sell recommendations can alert us to potential issues),
Unusually large share price moves, and spikes in volatility and trading volumes,
The presence of market or sector wide issues not captured by our quant process which are affecting market sentiment (eg regulatory risks).
Nevertheless, it’s still a time intensive process which is best suited to one region.
Focussing on one region also facilitates intra-day trading and doing well over 100 daily trades. Breadth is a very important component of our investment process which applies not just to the number of positions in the portfolio but also the number of daily trades. It’s an exhaustive process which would be very difficult to implement over a time period spanning multiple regions (eg Asian and European trading hours).
The question then becomes which region and markets should we invest in?
We can quickly and easily rule out the United States. There is too much quant money in the United States to effectively exploit factor driven pricing anomalies. This is reflected in our backtest results.
There is a relative paucity of quant factor money in Asia and Europe and our backtest results in both regions are strong.
So why do we invest in Asia rather than Europe?
There are several reasons:
Capacity
Asia has more than double Europe’s capacity. Once you move outside the main index names in Europe, liquidity rapidly declines. The following chart shows the median daily turnover of a universe of stocks ranked from 500 to 1,000 by market cap in Europe and Asia.
source: OQFM, FactSet
We believe that many investors underestimate the importance of capacity and the extent to which market impact costs can erode pre-cost alpha. As quants, we know how big an issue it is.
To illustrate: if we were run our investment process in Europe, capacity would be less than $350m.
Breadth
We need breadth to successfully implement our investment process. We take a “probability” rather than “perfection” approach to investing.
To understand why, consider the Fundamental Law of Active Management: IR = IC * sqrt(Breadth).
This law states that risk adjusted performance is a function of the manager’s skill (IC) and the number of independent investment decisions (Breadth).
The quant factors used in our systematic stock selection process have limited predictive power. We can improve the predictive power of our process by combing the factors in a smart way, but there is a limit to what can be achieved. Market anomalies tend to be subtle rather than conspicuous.
We can successfully compensate for the lack of predictive power by utilising large breadth. We want as many investment opportunities as possible.
Our stock universe is based on analyst coverage and Asia has far more stocks with analyst coverage than Europe, and slightly more than the North America.
Source: OQFM, FactSet
Stock knowledge
I started doing Asian quant factor research in the mid-1990s. I then moved to the buy side and launched an Asian quant hedge fund using a discretionary multi-factor quant process in 2005. The European fund wasn’t launched until 2011.
Given this I have more experience with Asian stocks than European stocks. The same applies to Zicai and Wilson who joined the Macquarie team before we launched the European fund.
Detailed stock knowledge facilitates the successful implementation of our discretionary overlay making Asia the preferred choice.
Time zone
The OQ team is based in Hong Kong. We run an intra-day rebalancing process and start trading earning in the morning and finish at 5pm when Singapore closes. It’s an intensive process which would be difficult to implement during European trading hours which would necessitate working up until midnight Hong Kong time.
Market characteristics
Asian equity markets have various characteristics that favour regional specialism and the ability to generate alpha via a discretionary overlay:
A/H dual listings: Numerous Chinese stocks are listed in both Hong Kong and China. They’re not fungible and there are considerable relative share price differences which vary over time. Currently, the vast majority of ‘H’ shares trade at a significant discount to their ‘A’ share counterparts. Currently, there is no mechanism to close the discount, but the size of the discount is still relevant. For stocks with high payout ratios, the difference in dividend yields can be substantial. For stocks with distressed balance sheets, the ability to raise money relatively cheaply via the ‘A’ share listing is also pertinent (eg there are examples of ‘A’ share placements being made at a significant premium to the ‘H’ share price).
Large net cash holdings: Asia is unique in that there are a relatively large proportion of stocks with significant net cash holdings. This is particularly the case in Japan, but it also applies in the other markets in our stock universe (other than Australia). We systematically screen for stocks which have attractive earnings and cashflow profiles relative to their enterprise value. Our discretionary overlay then examines whether companies are proactively looking to use the cash to increase the shareholder yield.
Holding companies and companies with large external shareholdings: There are numerous holding companies in Asia and companies which have significant shareholdings in other listed companies. This can distort systematic value analysis. For example, stocks can look expensive based on valuation factors but look cheap on a sum-of-the-parts basis. To illustrate: Fila (081660 KS) has a holding in Acushnet (GOLF US) which is roughly equivalent to its enterprise value. Similarly, Singapore Telecom (ST SP) has stakes in regional telecoms companies which roughly equate to its enterprise value. There are many similar examples in Asia.
Currency sensitivities: Unlike Europe, Asian countries have their own currencies. And currency moves have a significant impact on relative and absolute share price moves. In particular, Japanese equities have high currency betas, higher than any other large developed equities market. This is complicated by the fact the Yen is seen as a safe haven and often moves sharply in the opposite direction to other regional currencies during periods of risk aversion. Understanding currency dynamics in Asia helps in the development of systematic models and the implementation of our discretionary overlay.
Dividend-ex-date quirks: In Korea, stocks go ex-dividend before the dividends are announced and the actual amount isn’t known for months after the ex-date. In Japan, most stock go ex-dividend on the same date. In Taiwan, stocks often pay large stock dividends (which are basically bonus issues) which have adverse withholding tax implications for offshore investors. In Australia, there is an imputation system which affect the after-tax value of dividends to both onshore and offshore investors (offshore investors don’t have to pay withholding tax on fully franked dividends).
Asia is clearly superior to Europe for our style of investing. We will continue to invest solely in Asian equities.
Comments