Fund categorisations: confusing segments and style
The Indian media has been discussing SEBI’s circular on categorisation and rationalisation of mutual funds ever since it came out in October 2017. I had written about the need for better fund classifications last year, so I didn’t comment further. But the news that AMFI has asked SEBI to reconsider has prompted me to share my views a bit more openly.
I believe the industry is confusing segments and style. It’s also confusing numbers with consistency. We need to set standards that ensure funds are true to label but allow variety of styles.
Let me explain.
Segment is about what you invest; style is about why
Investments into assets can either be as debt or equity. We then create ‘asset classes’ by term, underlying exposure, liquidity etc to group assets that have similar risk/return characteristics. Most text books divide asset classes into cash, fixed interest, real estate, commodities and equities. There are also alternative asset classes that have limited liquidity such as private equity and infrastructure.
We also classify by geography into domestic and international/global because there are practical issues such as capital controls, availability of mutual funds etc. Hence, Indian equity is a separate asset class from global equity. For an American investor though, US equity is separate from international. (Global includes whole world, while international is global minus domestic; this distinction matters when the domestic has significant market cap share).
We can keep sub-dividing by quality, size etc. Government bonds are separate from investment corporate bonds, which in turn are separate from high yield or junk bonds. Similarly, blue chip stocks will have different risk/return characteristics from small cap. But some call these ‘segments’ rather than ‘asset classes’. Why? Market practice.
Asset classes and segments are not ‘scientific’ or set in stone – they are based on generally accepted market practice. Ideally, the industry should agree on this, failing which the regulator steps in.
Once the investor has decided what to invest in, she select the mutual fund in that asset class and segment, and then leaves it to the fund managers’ discretion on how to invest. The fund manager can use a top-down or a bottom-up approach. He can use a quantitative filter or assess company management on gut feel. He can buy into Jeff Bezos’ or Elon Musk’s vision of growth, or stick to stocks that have profitable cash flows. He can follow the cigar-butt value approach that Ben Graham wrote about, or the moat approach that Warren Buffett talks about.
There are as many ways of picking stocks as there are investors. And investors evolve as they get more experience, as markets change, as the world changes. That’s his thought process, his philosophy, his style. (We will go into nuances some other time).
Some people point to the commonality of stocks in mutual fund portfolios as evidence that fund managers don’t have style differentiation. They miss the point. Two fund managers or investors can use very different styles and end up holding the same stocks at given point in time. As a researcher, I ask the fund manager why they bought or sold that stock. When looking for evidence of style, I check when and what price they bought and sold the stock.
We need consistency of segments, variety of styles
Let’s turn to the issue of there being too many funds in the market causing confusion for investors.
The number of funds is not the issue; the lack of consistency in segments is. The segments are not clearly defined in terms of – what constitutes that segment, what the appropriate benchmark is, and how to handle changes. Hence, a large cap fund by one asset management company (AMC) could be investing in largest 50 stocks while another defines large cap as the largest 200 stocks. Rating houses classify them into different segments. Worse, the funds get moved around, for example from multi-cap to large cap, back to multi-cap segment, depending on either the fund manager changing the portfolio or market changes. I’ve been pointing this out to the industry for years (in my Mercer days), but it takes consensus or an independent arbiter to resolve such inconsistencies.
Style, on the other hand, is hard to define. As I pointed out, Warren Buffett defines value differently from how his own mentor, Ben Graham defined it. Indeed, Buffett believes all investing is value investing, in that every investor wants to pay less for what he perceives future value. We can’t put value in neat little style boxes. More importantly, we don’t need to.
We have to assess the fund managers’ ability to add value. How he does it, frankly, is his problem. If he uses a value style, it’s his choice. If he uses a growth style, that’s his choice too. I have to understand his style only to understand when (not if) it’s likely to underperform. I can then choose whether I can withstand his style-related underperformance, or whether I want to smooth out my portfolio performance by adding one or more funds with other styles.
Good fund managers have one style. It’s hard enough to find one way of adding value; it’s impossible for the same person to find multiple ways to add value (I am sure, unless someone proves otherwise). It’s like saying it’s possible to live life using multiple, conflicting values. So AMCs start out with one style that their lead fund manager has come up with.
The problem starts when that fund does so well that it grows to such a big size that the style no longer works as well. So the AMC launches new funds with different fund managers and styles, with the intention of diversifying their business risk and enhancing their ‘capacity’. Large AMCs then claim to be able to manage multiple styles without any issues. AMC’s research processes are organised around a particular style. While it’s possible to use the same research for multiple styles, it’s hard. The capacity argument is a bit misleading too as any limits to stock ownership are at the AMC level, not fund level.
In this context, the regulator stating AMCs can only have one value or contra fund is interesting. Value and contra are styles, hence hard to define – so they shouldn’t be segments. Defining them as segments impinges on larger AMCs to manage their business and capacity. It can also be confusing as smaller AMCs using a value style can now be put into two different segments.
Portfolio construction is not just about number of funds
Why do we have so many segments and styles to start with? It’s to facilitate portfolio construction. If I only had one goal, say to maximise wealth when I retire, I could just buy one fund (whether value or index, depending on your investment philosophy) and never touch it until I retired. The problem is I am human – I sometimes panic when my fund values drop. I also have goals other than retirement, so I will need to liquidate some of my investments along the way. Fund managers are also human – they sometimes don’t evolve with the market or lose the plot, and underperform on an ongoing basis. AMC staff are also human – sometimes they make mistakes such that there are operational blow-ups.
To manage all of these human tendencies, I decide to diversify by investing in a portfolio of funds. However, diversifying across a number of funds won’t help if they behave the same way. The funds should behave differently, but more importantly, they should behave as per expectations. Just like an equity fund shouldn’t invest in bonds, a large cap fund shouldn’t invest in small caps.
Once the lines are drawn and expectations set, the fund should stay within that line. Once fund manager states their style as an explanation of how they intend to add value, they should stick to it, or inform investors. Only if funds perform in line with expectations, can investors build portfolios appropriately. Until then, investors seem to settle on a random number of funds that they should have in their portfolio without any rationale whatsoever.
It’s great that the regulator has ignited the debate. Ideally, these debates should happen in the industry, at conferences and in trade media, before the regulator formulates regulations. Let’s hope we do so in the future.