DSP Mutual Fund has launched a new fund offer (NFO) for DSP Quant Fund. The new scheme proposes to select stocks based on pre-determined rules and eliminate fund manager bias in the stock selection process. The NFO, which was launched on 20 May, will remain open till 3 June. Since the fund is open-ended, you can also invest after the NFO closes. The fund will be benchmarked to BSE 200 (Total Returns Index). “For any decision making - whether about what food to order, or what investment to make, biases often make us compromise our primary beliefs. Hence, we have tried to create this fund with minimal bias," said Anil Ghelani, fund manager, DSP Quant Fund.

How it will work

The fund will eliminate companies which are highly leveraged, highly volatile, where management incentives are misaligned or where there is poor earnings quality. The fund will also apply quality, growth and value parameters such as earnings growth, return on equity (RoE) and dividend yield to narrow down the list. The fund will restrict exposures to single stocks and single sectors at 10% of assets.

It will follow semi-annual rebalancing to minimize turnover (which increases costs and eats into returns).

The fund house has back-tested its model. The results indicate that the fund will do well in some types of markets and underperform in others (for example, it will underperform in overheating markets).

Such schemes usually come in for three types of critiques. First, that the model works only in back-tests. In other words, since markets are inherently unpredictable, there is no reason why a model which has worked in the past should also work in the future. The AMC’s response is that the model has been tested over long periods.

Second, there is the criticism that the model is a data mining exercise. What this means is that the model may have settled upon stock filters which have no real significance, simply because they fit the data. To this, the fund house has responded that the factors on which the model is based have been tested across geographies and time horizons.

Third, there is the argument that quant means high-frequency trading. Quant funds, especially quant hedge funds, tend to use algorithms and high-frequency trading to generate returns. This often ends up simply pushing up costs through brokerage and securities transaction tax without outperforming a traditional mutual fund. Responding to this, DSP AMC said that the portfolio will be rebalanced on a semi-annual basis to avoid transaction costs from excessive churning. The fund’s model shows it to be overweight in IT, materials, consumer staples and consumer discretionary and underweight on energy and utilities.

The AMC has also indicated that the fund would have lower expenses than its actively managed peers. “The fund is in a sweet spot between active and passive. Hence it would come at a lower cost to the investor since the fees charged would be just slightly higher than a plain-vanilla passive index fund," said Ghelani."

Experts are divided on the merits of the fund. “Outperformance of active funds over rules-based passive funds is falling, although it will take another five to 10 years for it to completely play out. Investors with a long time-frame should build an allocation to quant in their portfolio. This should come out of their large- and multicap allocation," said Samant Sikka, founder, Sqrrl, an online mutual fund investment platform.

However, other experts took a more cautious tone. “I like the idea of a quant fund which hopefully reduces subjectivity. However, I would wait for the fund to show some performance before I recommend it," said Suresh Sadgopan, founder, Ladder 7 Financial Advisories.

Mint take

The idea of removing human bias is an appealing one. However, an experienced manager can spot issues in a company that a rules-based system may not. Quant funds can be quite dissimilar to one another, but one example of an existing fund of this nature, Reliance Quant Fund, has failed to beat its benchmark, the BSE 200 (Total Returns Index) over the past three and five years.

Quant funds are sold on the premise that they contain losses in bear markets while slightly underperforming in strong bull markets. The record on this is mixed in the case of Reliance Quant Fund.

Last but not least, the sectoral allocation of DSP Quant Fund, as indicated in its presentation, seems to be tilted towards relatively richly valued sectors. Whether or not it can outperform actively managed funds is still an unanswered question. Investors should wait for the fund to establish a track record before putting their money in it.