JioBlackRock Mutual Fund is using a rules-based model to decide how much of its Balanced Advantage Fund should be invested in equities and which stocks should make the cut. The framework combines global risk sentiment with four local signals before applying six factors to a universe of 250 companies.
The model can move the fund’s equity allocation across predefined brackets ranging from 20% to 90%. It then selects stocks by weighing expected returns against risk, liquidity and trading costs.
“Being in India, we all look at local signals, so a global signal is not so common — that is a big differentiator for us,” said Rishi Kohli, chief investment officer at JioBlackRock Mutual Fund.
The global signal is risk sentiment, which captures the impact of factors such as geopolitics, crude prices and bond yields. The local indicators cover valuation, technicals, local risk sentiment and macroeconomic conditions.
“What is important is that there is everything there from valuation to technicals to local risk sentiment to macro,” Kohli said. “When you combine the global risk sentiment with these four local buckets, you are virtually covering everything that is important.”
The first stage of the model determines how much equity exposure the fund should carry. Each signal is assigned a weight, and the combined score determines the appropriate equity bracket.
The model could, for example, recommend a 50% allocation to equities. Kohli said the paper portfolio had been in the 40% to 60% range over the past six months, moving between slightly overweight and slightly underweight positions depending on market conditions.
“A 50% means it is neither bullish nor bearish, it is somewhere in between,” he said.
The technical indicators used by the model go beyond momentum. They also include reversal and breadth indicators, which help assess the health of the market and the breadth of participation.
The valuation framework uses traditional measures such as price-to-earnings and price-to-book ratios, but also gives greater weight to the earnings-bond yield. Kohli said the measure was more relevant for an asset-allocation model because it compares the earnings yield available from equities with bond yields.
Local risk sentiment is based on news sentiment. The model aggregates news related to individual companies and the broader economy to assess whether sentiment is improving or deteriorating.
Macroeconomic inputs include gross domestic product growth, inflation and interest rates. These factors can affect equity and bond markets differently across economic cycles.
The tax constraint
A Balanced Advantage Fund needs to maintain a minimum level of equity exposure to qualify for equity taxation. Kohli said that when the model recommends an equity allocation below 65%, the difference can be met through cash-futures arbitrage.
If the model recommends 50% equities, for instance, 35% of the portfolio could be invested in fixed income and 15% in cash-futures arbitrage. This would take the fund’s equity classification to 65% for tax purposes, while maintaining the economic characteristics of a 50% allocation to equities.
“That is not a differentiating factor, but just to make it very clear as to how we are achieving the 65%,” Kohli said.
Six factors determine stock selection
Once the model establishes the equity allocation, it selects stocks from the top 250 companies. The universe currently represents the large-cap and large-and-midcap segment.
JioBlackRock is not passively investing in the index. Instead, the model evaluates the 250 stocks using six factors: momentum, value, low volatility, quality, fundamental momentum and sentiment.
Fundamental momentum covers measures such as sales growth and earnings growth. Sentiment combines news sentiment with analyst sentiment, which Kohli described as a separate source of potential alpha.
The model begins with an equal-weighted approach, but adjusts factor weights depending on their effectiveness, expected churn and practical portfolio considerations. Each factor remains present in the framework because timing a rotation from one factor to another can be difficult.
“Global research has shown it is very difficult to time factors as such,” Kohli said. “What the best funds globally also do is they will have all the factors, and then maybe the weightages could change slightly.”
Aladdin weighs alpha, risk and cost
The model does not select stocks solely on the basis of their alpha scores. Each company also receives a risk score that captures volatility, factor exposures and sector concentrations.
The framework also accounts for the cost of implementing trades. A large order in a smallcap stock can create greater market impact than the same order in a largecap company. Participation rates and liquidity can affect the final price at which a portfolio manager is able to transact.
JioBlackRock uses BlackRock’s Aladdin platform to combine these variables. The optimisation engine evaluates alpha, risk and cost before determining the final stocks and their portfolio weights.
“So what we have on our Aladdin platform, on our Aladdin engine, is an optimisation which takes the alpha, risk and cost into account and then throws out the stocks and the weightages that will come out,” Kohli said.
The resulting portfolio is expected to hold between 50 and 75 stocks, with a typical range of 60 to 70. That makes it less concentrated than a focused portfolio, but less diversified than a broad flexicap strategy.
For Kohli, the objective is not to identify a single winning signal or factor. It is to combine multiple sources of information and make the portfolio-construction process more consistent.
“The alpha score means the best expected performance after taking these kinds of factors into account,” he said. “For us, it is a mix of all of these factors, where we are saying that it makes it more stable and a little more consistent.”
(Disclaimer: Recommendations, suggestions, views and opinions given by the experts are their own. These do not represent the views of The Economic Times)
