Framework for evaluating Mutual Funds – A look at DSP Quant Fund

The C.U.R.R.E.N.T framework

1. C – Communication
   • Does the fund manager communicate regularly
   • If the fund starts underperforming, will they let you know in plain English as to why this is happening
2. U – Understanding
   • Do you understand the investment process, portfolio construction strategy, and the logic behind the thought process?
   • What is the turnover in the fund?
3. R – Risk (using the NAVY framework discussed earlier here)
   • Normal Volatility
   • Abnormal Volatility
   • Valuations
   • Yearly drawdowns
4. R – Returns
   • Rolling Returns – 3Y & 5Y – % of times above 10%, 12%, 15%, Min return and Max return
   • Returns across a market cycle
5. E – Expense Ratio & Taxation
6. N – Negative return environment
   • All strategies underperform at some point in time. Under what conditions will the portfolio underperform?
7. T – Team
   • Fund Manager Track Record
   • Experience

I will take the example of a recently launched fund called DSP Quant fund to explain my framework.

DSP Quant Fund – Putting the fund to test

Let us start with

Check 1: Understanding

This is a quant fund where there is no fund manager and instead the portfolio is completely run based on pre-defined rules

How does it work?

Beginning with the universe of companies from the S&P BSE 200, the model builds the portfolio using a 3-step process.

#1: ELIMINATES companies with value diminishing components

• Exclude companies with high leverage (Debt to Equity > 1.5)
• Exclude companies which have high price volatility (Beta >1.25)
• Exclude companies which are inefficient capital allocators and do not work with the sole aim of shareholder wealth maximization (Govt. owned and state-owned enterprises)
• Exclude companies with poor earnings quality (as defined by our earnings quality screeners)

After applying the exclusion criteria for recent backtests, the universe is reduced to about 80-100 companies.

#2: SELECTS good companies with durable sources of potential out performance

• Percentile score assigned for each company across selected factors like Quality (ROE, Earnings growth variability), Growth (Estimated consensus EPS growth) and Value (Dividend Yield & Free Cash Flow Yield) which is combined into an aggregate score for relative company percentile ranking (equally weighted for each factor).
• Include for consideration only top-ranked companies (highest aggregate score) which constitute 50% of BSE 200 index by weight.
• This further reduces the stocks that will be considered for inclusion in the portfolio to about 40-50 stocks in recent rebalances as per back-tests.

#3: OPTIMIZES portfolio by allocating weights to companies

Each selected stock is weighted appropriately to reduce stock & sector concentration and liquidity risk.

The portfolio is re-balanced semiannually (March and September).

This is pretty much what any fund manager would attempt to do. Everyone starts with a wide universe and eliminates stocks which according to their framework would be high risk and have lower return potential.

Intuitively it makes sense as all they are trying to do is eliminate 1)Weak balance sheets (high debt) 2) Stocks with Suspect Accounting 3) high volatile stocks 4) PSU companies (perceived to be weak management)

Then they try to strike a balance across the factors – Value, Growth and Quality.

This is in alignment with the factors which have worked well globally – Value, Quality, Growth, Momentum, Low Volatility, Size etc. They have not considered momentum factor and size. Low volatility is indirectly addressed via eliminating high volatility stocks.

What is the churn of the portfolio?

The re-balancing is done only once every six months, and the turnover is in the 40-50% per annum range. This means 40-50% of the portfolio will approximately change over a year. This is neither too high nor too low.

Overall, the investment process is pretty straightforward and intuitively makes a lot of sense. While I don’t understand the optimization part completely, the overall intent is to check for stock, sector concentration and tries to ensure adequate representation of all three factors (value, growth, quality)

Check 2: Returns

The back tested performance has comfortably beaten the Nifty, Nifty Next 50 and large cap category over the last 13 years.

Let us check the rolling returns data

The model has increased the odds of a >12% return over both 5 and 7 year periods.

Market Cycle Returns

Let us check the returns across the across an entire market cycle – from the peak of last cycle (Jan-2008) peak till date

As seen above, over the complete market cycle, the DSP Quant model has comfortably outperformed, both passive indices and active funds.

Check 3: Risk

Let us apply the N.A.V.Y framework

Normal Volatility

On a 6 month basis, when you review the fund, we need to first set the right expectation on what should be considered as normal ups and downs. So I will be removing the extreme events on both sides (positive and negative).

The DSP Quant model, has a 95% probability based on history, that its returns will be in the range of -30% to 52% over the next 6 months. While we cannot predict where exactly it will fall in this range, when you are reviewing the performance every 6 months, returns anywhere in this range would be considered normal.

The 6 month normal downside is relatively lower than other passive benchmarks.

Abnormal Volatility

As seen above, the drawdowns are relatively lower and the recovery is also faster. But the key point is that in the worst-case investors must be ok to take the temporary pain of a 54% drawdown (going by history).

Valuations

If you noticed, both on a PE and PB parameters, the fund is trading close to its 2008 peak levels. Assuming mean reversion over the next 5-7 years, this means for those investing at these levels, investors will only be able to partially translate earnings growth into returns.

Hence earnings growth remains the key for future returns as valuation upside from these levels looks difficult.

Yearly Drawdown

In 2008, the DSP Quant Model was down -46% vs the -24 to -32% fall in passive indices -51% to -63% (BSE 200, Nifty 50, Nifty Next 50)

In 2011, the DSP Quant model was down -15% vs the -24 to -32% fall in passive indices (BSE 200, Nifty 50, Nifty Next 50)

The fund given its tilt towards quality has fallen lesser compared benchmarks indicating relatively lower risk.

The average intra-year drawdown has been around 13% (ignoring 2008) vs 15-19% for other indices

Thus overall the fund has demonstrated lower downside and volatility compared to other indices. The current above-average-valuations remains the key risk.

Check 4: Team

Now this is a quant fund and hence the fund manager is irrelevant. The fund would simply follow the rules slavishly.

Check 5: Expense Ratio and Taxation

The direct fund expense ratio is expected to be ~0.4%. This is a significant advantage compared to active funds at 1%+ as the lower cost over the long run adds significant return advantage to the portfolio due to compounding effects.

The taxation is the same as any other normal passive or active equity fund.

Check 6: Negative Return Environment

The fund may underperform its benchmark during periods of

• Sentiment driven rallies / Market Euphoria (Not backed by fundamentals) Example: 2007 CommoditySuper-cycle peak and 2014 Change in Govt. regime

• Market reactions based on actual or expected changes in policy/regulation or events. Example: PSUBank recap of Oct 2017

• The fund may not be able to fully capture ‘hope trades’ or ‘turnaround stories’ where actual historical numbers are poor but market is pricing in a sharp future improvement

Check 7: Communication

Historically the AMC has done communicated really well, both in terms of transparency and timely updates.

The communication for this fund was one of the best I have seen in recent times.

Check the links to get a glimpse of all the product related materials

• Product Presentation – Link
• FAQ-1 – Link
• FAQ-2 – Link
• Primer on Factor Investing – Link
• Common mistakes in backtesting and how to avoid them – Link

The Eighty Twenty Investor View

Pros

1. Transparency: Clearly defined intuitive process
2. Low cost
3. Rules based & Systematic
4. Improved odds of higher returns vs benchmark over a 5 year period (going by the backtests)
5. Relatively lower downside and volatility
6. No problem of fund manager changes and style drift

Concerns

1. Valuation of the model is at its highest – at the end of the day, future returns is simply the sum of valuation movement + earnings growth + dividend yield. Since valuations always mean revert over longer horizons (which means they will go down), the returns going forward will be a partial capture of earnings growth. Earnings growth will remain the key going forward.
2. As someone who has done backtests for a living, I am always skeptical of backtests. Real-life always has its fair share of surprises!

I think this is an interesting fund and soon I expect other AMCs to follow. In a year or two we will have a lot more rule-based funds.

As options increase and track records get built, I believe future portfolios will have a significant portion built via low cost, passive rule based strategies.

As of now, I will continue to watch this fund on how it translates into real life performance. Barring valuation concerns in the short run, I believe it makes for an interesting offering in the longer run.

Let us wait and watch 🙂

Future Research Required: Rising valuations, above their historical normal levels, can sometimes artificially inflate past performance and reduce the future return prospects of a rules based strategy. Hence removing out the effect of changing valuations on past returns results in a more reliable estimate of a strategy’s true excess return producing ability.  Since I didn’t have the valuation data for the model I wasn’t able to do it. I shall reserve it for future research.

Before you get lost in the fund specifics, the overall intent actually was to present the new C.U.R.R.E.N.T framework for evaluation of equity funds. I hope you found it useful and do share your feedback on how to improve this further.

This post was a guest contribution by Arun, who is popularly known as the eighty – twenty investor. You can follow him on Twitter here. He runs a blog which can be accessed here.

This post was originally published here.

Read more about my story here.

Subscribe to my newsletter here. It goes out once a month.

Get the latest post alerts directly on your Telegram.