Are “debt like” mutual funds better than pure debt funds?

Many thanks to Sayan Sircar for helping me understand many concepts that led me to eventually write this blog post.

Ever since the Indian government decided to tax debt mutual fund gains at the slab rate, I have been struggling to decide on a course of action. My initial intuition was that we should invest in the chosen assets and not worry about taxes. Then I changed my mind and I was attracted towards “debt like” funds that invest a significant portion of the capital in arbitrage.

But I had a constant nagging in the back of my mind because:

1. Arbitrage is constant work (for the fund management team) while debt funds are not as much work. I like the passive style of investing where “buy and hold” is the norm.
2. Arbitrage may not make as much money in bear markets (since futures may not sell at a good premium).

Arbitrage behaves differently than bonds, and I don’t exactly want arbitrage in my long-term portfolio. Until today, this was mostly intuition, but now I have some data to back it up. To make sense of that data, we must first understand correlation.

A primer on correlation

Correlation quantifies how 2 different investment assets behave relative to each other. A correlation of 0 means uncorrelated. When one asset’s movement doesn’t affect the other, we say they are uncorrelated. Cash is uncorrelated with all other asset classes.

Positive correlation means the assets move in the same direction; negative correlation means the assets move in opposite directions. We don’t want the correlation to be close to +1.0, which means the 2 asset classes behave mostly the same. That means there is no effective diversification.

A correlation close to -1.0 means the 2 asset classes cancel each other’s appreciation and we’d often end up with net zero return. This may sound unappealing, but rebalancing such a portfolio can be very rewarding. (In other words, such portfolios offer high rebalancing bonuses.)

Correlation between a few Indian mutual fund categories

I tried to get data for a few decades, but my attempts failed. ValueResearch provides calendar year returns for the last 9 years (2016 through 2024) at the fund category level. 9 years is not a long time period, but this is probably okay since this period includes both low and high interest rate markets. One interest rate cycle is better than zero! 😅

The correlation between mutual fund categories are shown in the following table.

Large cap : Arbitrage	-0.0312
Flexicap : Arbitrage	-0.0114
Large cap : Liquid	-0.2562
Flexicap : Liquid	-0.2593
Large cap : Short duration	-0.1700
Flexicap : Short duration	-0.1264
Large cap : 10-year gilts	-0.3884
Flexicap : 10-year gilts	-0.3740
Large cap : Dynamic bond	-0.3593
Flexicap : Dynamic bond	-0.3017

I had the intuition that arbitrage was similar to cash. The data confirms my intuition since arbitrage has a correlation (with equity) very close to 0. Shorter duration debt funds have correlation closer to -0.25 while longer duration debt funds have correlation closer to -0.4. (The Short Duration category falling between liquid and arbitrage is probably an anomaly caused by the small sample size of data we are working with.)

This is the metric I had to see to convince myself that pure bond funds are a better asset class to mix with equity. The lower correlation with equity means that an equity+bonds portfolio will have lower drawdowns than an equity+arbitrage portfolio.

My recent analysis quantified that the difference in post-tax growth rate is in the range of 0.5% to 1.5% annually. The difference in growth rate reduces as the holding period gets longer and longer. If we assume that a dynamic bond fund returns 0.5% more on average than an arbitrage+bond fund, then the net corpus won’t be very different between the 2 options.

Conclusion

1. Bonds, especially longer duration bonds, are a good asset to mix with equity. While arbitrage behaves like cash, bonds can provide meaningful downside protection (and provide meaningful rebalancing bonus).
2. For holding periods of 2 or more decades, the post-tax growth rate of bond funds is not too far from the post-tax growth rate of arbitrage+bond funds (source).
3. If we hold longer duration bond funds (such as dynamic bond funds), it is not unreasonable to expect that they’d yield 0.5% more return than arbitrage+bond funds. This further diminishes the difference in post-tax return.

The impact of “taxed at slab” — Take 3

I had quantified the difference between 30% tax and 12.5% tax—i.e., tax payable for bond fund gains vs tax payable for arbitrage+bond fund gains. The result of that previous analysis tilted in favour of paying less tax. The longer the holding period, the higher the impact of tax.

In a way, that analysis simply quantified what was already obvious. It is a well known fact that the gain (or value appreciation) increases exponentially as the holding period increases. The tax difference is flat 17.5% (30% – 12.5% = 17.5%). A flat percentage on a larger number will of course be larger than the same percentage on a smaller number. There wasn’t anything new to learn there other than quantifying what we can already know intuitively.

Now I want to measure the impact of tax differently, this time controlling for the exponential growth caused by longer holding periods. What if we calculated the CAGR of post-tax returns? That is, use a fixed pre-tax growth rate, compute the final corpus value, deduct estimated tax, and calculate the CAGR from original corpus to post-tax corpus.

When I do this, 2 things become noticeable:

1. The difference falls roughly in the 0.5% to 1.5% growth rate.
2. The longer the holding period, the lesser the impact on the growth rate. When I ran my previous analysis (i.e., take 2), this is what I expected to see, but I saw the opposite. It’s good to finally see data that validates my intuition.

Each cell in the following tables shows the post-tax growth rate. For example, look at the 6.5% row in the “3 years” table. This row says that if the pre-tax growth is 6.5%, the effective post-tax growth is 5.73% annualised when the tax rate is 12.5% and 4.63% if the tax rate is 30%. By paying higher tax, our investments (effectively) grow 1.1% more slowly each year.

For a holding period of 3 years, the difference in growth rate is 1.1%. If we extend the holding period to 10 years, the difference narrows to 0.96% (per annum). For 20 years, the gap reduces further to 0.78%.

But sensible readers would say that growth rates can go to hell. What really matters is the size of the final corpus, and hence, take 2 is a more useful analysis than this one. They’d be right, except the mutual funds that are taxed at 12.5% come with caveats.

An arbitrage+bond fund behaves differently than a pure bond fund. For some investors, the change in behaviour may not matter. But if the difference matters to you, you shouldn’t feel too bad about paying 30% tax. The impact of tax reduces as you hold the assets for longer and longer.

My conclusion is that long-term investors who plan on holding their investments for 2 or more decades can use this data to rationalise holding pure bond funds. This analysis doesn’t magically reduce the tax we pay, but it changes the narrative to convince oneself to choose a simpler investment asset (pure bond) over a more complex one (arbitrage+bonds).

Num of years ➡	3 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	3.96%	3.19%	0.77%
5.0%	4.40%	3.55%	0.85%
5.5%	4.84%	3.91%	0.93%
6.0%	5.29%	4.27%	1.01%
6.5%	5.73%	4.63%	1.10%
7.0%	6.17%	5.00%	1.18%
7.5%	6.62%	5.36%	1.26%
8.0%	7.06%	5.73%	1.34%
8.5%	7.51%	6.09%	1.42%
9.0%	7.96%	6.46%	1.50%

Num of years ➡	5 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	3.98%	3.23%	0.75%
5.0%	4.43%	3.60%	0.83%
5.5%	4.87%	3.97%	0.90%
6.0%	5.32%	4.34%	0.98%
6.5%	5.77%	4.72%	1.06%
7.0%	6.22%	5.09%	1.13%
7.5%	6.67%	5.47%	1.20%
8.0%	7.12%	5.85%	1.28%
8.5%	7.58%	6.23%	1.35%
9.0%	8.03%	6.61%	1.42%

Num of years ➡	10 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	4.03%	3.33%	0.70%
5.0%	4.48%	3.72%	0.77%
5.5%	4.94%	4.11%	0.83%
6.0%	5.40%	4.50%	0.90%
6.5%	5.86%	4.90%	0.96%
7.0%	6.32%	5.31%	1.02%
7.5%	6.79%	5.71%	1.08%
8.0%	7.25%	6.12%	1.13%
8.5%	7.72%	6.53%	1.19%
9.0%	8.19%	6.95%	1.24%

Num of years ➡	15 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	4.07%	3.41%	0.65%
5.0%	4.53%	3.82%	0.71%
5.5%	5.00%	4.23%	0.76%
6.0%	5.47%	4.65%	0.82%
6.5%	5.94%	5.07%	0.87%
7.0%	6.41%	5.50%	0.91%
7.5%	6.88%	5.92%	0.96%
8.0%	7.36%	6.36%	1.00%
8.5%	7.83%	6.79%	1.04%
9.0%	8.31%	7.23%	1.08%

Num of years ➡	20 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	4.10%	3.50%	0.61%
5.0%	4.58%	3.92%	0.66%
5.5%	5.05%	4.35%	0.70%
6.0%	5.52%	4.78%	0.74%
6.5%	6.00%	5.22%	0.78%
7.0%	6.48%	5.66%	0.82%
7.5%	6.96%	6.11%	0.85%
8.0%	7.44%	6.56%	0.88%
8.5%	7.93%	7.01%	0.91%
9.0%	8.41%	7.47%	0.94%

Num of years ➡	25 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	4.14%	3.57%	0.57%
5.0%	4.61%	4.01%	0.61%
5.5%	5.09%	4.45%	0.64%
6.0%	5.57%	4.90%	0.68%
6.5%	6.06%	5.35%	0.71%
7.0%	6.54%	5.81%	0.74%
7.5%	7.03%	6.27%	0.76%
8.0%	7.51%	6.73%	0.78%
8.5%	8.00%	7.20%	0.81%
9.0%	8.49%	7.66%	0.83%

Num of years ➡	30 years
Growth rate⬇ / Tax rate ➡	12.5%	30.0%	Growth rate Δ
4.5%	4.17%	3.64%	0.53%
5.0%	4.65%	4.09%	0.56%
5.5%	5.13%	4.54%	0.59%
6.0%	5.62%	5.00%	0.62%
6.5%	6.10%	5.46%	0.64%
7.0%	6.59%	5.93%	0.66%
7.5%	7.08%	6.40%	0.68%
8.0%	7.57%	6.87%	0.70%
8.5%	8.06%	7.35%	0.71%
9.0%	8.55%	7.83%	0.73%

When an abundance of information doesn’t help

Conflict of interest is a weird thing. It’s hard to detect and assess. Even when we can see a conflict of interest, it’s difficult to know what the conflict actually means.

When I used to be in the “anti-gold camp” (meaning, when I believed that gold should not be part of investment portfolios), I noticed that everyone that praised gold was also selling investment gold. That’s a clear conflict of interest. But just because someone sells gold, it doesn’t mean they have ulterior motives, right? On the contrary, someone who genuinely likes gold will of course sell it on their platform. (“Shocker! Milk seller thinks that milk is a good food!”)

On the other side of the marketplace we have asset managers boasting their “skin in the game.” Imagine you are asked to fly into enemy territory as a passenger on a fighter plane. The fighter pilot assures you, “I am flying right with you. I would genuinely try to keep us both alive.” The pilot isn’t lying. But taking that flight is a foolish move for most civilians. That the pilot takes the same risk as the passenger is irrelevant.

That means, making sound investment decisions boils down to knowing what you are doing. Investors need to understand the assets they invest in and decide for themselves whether they want to invest in an asset or not, how much they want to invest, etc. Fiduciary advisors can (and do) help, but the investor is the ultimate decision maker.

Academic research can potentially help, but acting based on academic research is also mired in complexity. Who funded the study? What are the researchers’ preconceived biases? Was this a “research” fabricated to prove a point, or was this done with genuine intellectual curiosity? How sound are the data and the methodologies used by the researchers? Most people cannot evaluate these. Even the conclusions of such research studies are beyond many ordinary people’s comprehension. As a result, ordinary people rely on someone else to simplify the study’s conclusion, and we are back to the same problem that we started with.

Thinking about all these reminds me of the phrase, “Water, water everywhere, but not a drop to drink.” An abundance of information is available to us, and yet, it is so hard to make sound investment decisions.

Photo by Pok Rie

Bollywood movies are like multi-asset portfolios

What is the composition of a typical Indian movie? Some drama or action, but often both. A few songs. A few fight sequences. Some comedy. It’s almost always a mix of everything. Leaving anything out is so uncommon that such a deviation will be talked about whenever such a movie comes out.

Why do movie makers do this? Why do they add everything to the movie even when some elements (such as comedy or fight sequences) are unnecessary for the story? Because movie making is a business, and the producers want to make money. Audiences may not like the hero’s face, but they may come in to see the heroine. Audiences may not like the songs, but they may come in for the comedy. Essentially, the producers are spreading their bets.

This is not very different from multi-asset investment portfolios. We add equity, bonds, gold, and maybe even cash. We add all that to make an “all weather” portfolio that can navigate any and all challenges that may arise in future.

If you are making a movie for one individual person to watch, you can just do what that person would enjoy. If you are making a movie for me, you can rip off songs and fights. Just give me the drama with beautiful imagery and good dialogues, and I am happy. But no one’s making a movie for just one kind of audience; they want to attract as many diverse kinds of people as they can.

Likewise, for any investment duration, there definitely is a single “winner” asset. If we put all our money in that one asset, we’ll come out with the maximum possible gain. Unfortunately, though, we cannot know that asset in advance; we can know it only after our investment duration has passed. So, in preparation for an uncertain future, we spread our bets into multiple asset classes.

Multi-asset portfolios are a compromise by definition. You could also argue that investors choose multi-asset portfolios only when they don’t have conviction in any asset class. I think Warren Buffett has an equity-only portfolio. But most of us aren’t Warren Buffett; multi-asset portfolios are the right choice for ordinary people like you and I.

A portfolio is a composite

No one picks up a bay leaf or a star anise or a cardamom to eat. No one would consider them tasty. However, try making a biryani without them, and you’ll know how critical these spices are to a dish like biryani. It’s cliched, but these spices add up to more than their sum in a dish like biryani.

Photo credit: Gourav Sarkar from Pexels

A diversified investment portfolio is similar. You may not quite like the individual assets on their own, but you may still hold them if an asset combination is more stable or enhances growth.

Back in 2021, I looked at gold in isolation and decided that it wasn’t a great asset to invest in. In 2024, I learnt that 2 risky assets can be combined to make a portfolio that’s less risky than either of those assets! I also learnt how incredible an asset gold is when combined with equity and bonds. These learnings have led to me to change my stance.

I still cannot embrace gold 100% because it’s a speculative asset. Gold investors make money only when other investors are willing to pay a higher price for the same gold. But it’s hard to ignore the stability and growth potential that a tinge of gold adds to a portfolio.

When I think of my portfolio as a composite (that has X% equity, Y% bonds, etc.), it becomes easier to have an allocation to gold—similar to adding bay leaves and star anise to a biryani. The final output, the composite, is more important than the individual ingredients that make up the composite.

Asset Allocation Vectors by Vecteezy

Thinking of portfolios as a cooked food item also makes it clear why regular rebalancing is very important. If your dish has too much salt or too little water, you don’t just ignore it. You intervene. You bring the ingredients back to an acceptable proportion. You don’t let one ingredient spoil the entire dish. The same is true for portfolios: you don’t let one asset become too heavy or too light, lest that one asset spoils the balance of the entire portfolio.