Thursday, September 30, 2010

Does sector investing fit into asset allocation? How?

The short answer is that it does, if you have time and energy to devote to personal finance. But if you are like the majority of people who do not have time for finance, and are minimalist and ad-hoc in their efforts, you would be better off without sector investing.
There are two reasons we say this:
1. Sector investing calls for deeper research than investing in diversified or index funds. After all, if you note the five-year performance of most diversified funds, they will differ from each other by only a few percentage points. This would not be the case with sector investing - you can get it spot on, or go very wrong.
2. Sector outlook can change more quickly than the view on Indian markets as a whole. For instance, you can continue investing in an index fund for decades, without worrying about ups and downs of the markets. But it is rare that a single sector will outperform for longer than 2-3 years. Therefore you need to have consistency of monitoring, to be able to enter and exit in time.
Before you continue, decide for yourself whether or not sector investing is for you. There is nothing intrinsically inferior about sticking to index investing - an ``invest it, shut it, forget it`` philosophy. Read on only if you have the commitment to do more.
Sectors in the equity asset class
Equity as an asset class is particularly interesting, in that it gives you a lot of choice within. This can also ruin you - for example by encouraging trading (i.e. frequent churning that enriches only the broker). But if used well, it can give you the thrill of market ups and downs, and of getting your analysis and foresight proved right.
Sector investing may be one such interesting offshoot of equity. It should never exceed about 30% of your equity portfolio (the rest being in diversified stocks or funds). You don`t want to have too many sector calls at once - since that is same as diversified investing! Two or three well researched calls at any point of time should do.
A time horizon of 1-3 years should be your target. Less than a year and you are getting into trading territory, not into strategic sector investing. Greater than three years, and it is likely that your calls haven`t really gone right - other sectors may have outperformed in the interim.
The approach
While we cannot reduce the art of investing into a step-by-step algorithm, we can lay down some principles and best practices based on experience, to aid your approach. Some of the best practices are:
> Unless you have some very specific and deep information, avoid sectors that have already outperformed the index in the last couple of years. An encore is unlikely.
> Avoid the dual risk of identifying a sector and a stock within it. One way to start is to only choose a sector, and invest in a sector mutual fund. That way, you are spared of individual stock research.
> If a sector is already hot news in the media and among `experts`, it is probably best to avoid it. Such sectors may have already run up in anticipation, so future potential may be limited.
> Ideally, derive your sector view from reading a general sector report from somewhere other than an equity research firm. Such reports tend to be more balanced, have long-term views, and are free of sensationalism. That is, you are keeping your sector view based on macro-trends and economic fundamentals.
> Measure your performance once a year, always comparing your sector with the Nifty index. Only an outperformance here means that you have got it right; not just a positive return.
> As mentioned above, make not more than 2-3 picks; each with a 1-3 year horizon. Keep this within about 30% of your allocation to equity. In each pick, be clear what the drivers of value for the sectors are (e.g. more foreign investment, greater consumer demand, government support, cheaper raw materials, etc). This way, you can monitor and decide whether things are going the way you anticipated.
With some effort and experience, you should find this an interesting experience. And make some useful above-normal returns too!

No comments: