Revathi Asset Management

Portfolio construction and fundamental screening

When I first started studying the markets, I was obsessed with finding the right metrics that could build the best portfolio over a long period. Essentially, the idea was to algorithmically manage a portfolio, choosing and rebalancing among stocks just by looking at screeners of fundamental ratios. If you read popular books on this style of investing, you’ll find that high ROE/ROCE, high margins, predictable growth, and predictable cashflows make the top of the list. Foreign portfolio investors especially follow such methodology very closely when investing in India.

The logic then follows that such companies should outperform those that have more volatile cashflows/margins, less predictable growth, and lower ROE/ROCEs. One would expect, given that such institutions manage the overwhelming majority of investments, these stable, predictable companies would give better risk-adjusted returns (even if not higher absolute returns) than their “lower quality” peers. Fast-moving consumer goods (FMCG) is one such sector in India that is a favourite amongst most foreign portfolio investors.

Personally, however, I always had a fascination for real estate development companies. Surely, I thought, in a land-constrained, over-populated country like India, real estate companies should do well. But I was told that they were simply too volatile, they were subject to extreme booms/busts that made them uninvestable, and they had destroyed investor wealth. They were also linked to major collapses/frauds in the past in the housing and infra sector, as we had seen earlier with IL&FS. I was told that no institutional or foreign portfolio manager would be caught dead holding real estate development companies in India.

Let us now compare two FMCG darlings (NESTLEIND and HINDUNILVR) vs two well-known real estate development companies (GODREJPROP and DLF). All 4 companies were amongst the top established companies in their sectors at the start of the period. Furthermore, the chosen FMCG stocks are cornerstones amongst foreign portfolios. Let us start by comparing sales, which should be highly predictable as per common advice.

We see that the FMCG companies fit the bill much better, showing almost a linear extrapolation in their sales trajectory. The real estate development companies, on the other hand, are extremely volatile and have not even shown any growth in absolute terms over 8 years.

Margins, cashflows, and ROEs

Let us now look at operating margins, where we want to see some stability. Looking at the FMCG companies, we see why foreign investors love them: their margins barely move and when they do, it’s to the upside. The real estate development companies on the other hand are all over the place. DLF at least appears to have high margins but it also has high volatility.

Let’s take it one step further to look at the operating profits. These should at least show some growth over time, given that the Indian GDP has grown from USD 2.29 Tn to USD 3.67 Tn from 2016 to 2024, i.e. at a CAGR Of 6.14%. Due to stable margins and predictable, trending sales for the FMCG companies, we see that their operating profits are also very stable and predictable. For the real estate development companies, however, as we are now starting to suspect, the operating profits are volatile. Between them, GODREJPROP looks to be the worse of the two with both lower margins and lower sales.

Looking at the CAGR in cashflows from operations (CFO), we see almost abysmal levels of growth in the CFO for the real estate development companies. Note that to keep up with nominal GDP growth from 2016 to 2024, CFOs should grow at least 6.14% a year. We see that the FMCG companies are performing well in this regard. DLF slightly lags GDP growth, but GODREJPROP shows a CAGR in CFO of 0.05% over 10 years.

Lastly, let’s compare the ROCEs and ROAs. As an investor, surely you should consider the return on capital employed or on assets to see how efficient the business is. Again, the real estate companies don’t make the cut. By any metric, the FMCG companies are significantly better companies to invest in and it makes perfect sense why investor behaviourtoward the two sectors differs.

Share price performance

Let’s finally compare the stock price performance of the 4 companies on absolute and risk-adjusted terms to confirm our analysis. Looking at the NAVs, we find something highly unexpected.

GODREJPROP has been the absolute star performer amongst the 4, with DLF as the clear second. Overall, the real estate development companies have outperformed the FMCG companies.

When we study the performance stats, we see that not only do the real estate companies generate significantly higher absolute returns, but they also perform much better on a risk-adjusted basis. The max drawdowns are twice as high for the real estate companies as for the FCMG stocks, but despite this GODREJPROP still has the same calmar ratio as the better of the two FMCG companies. On the volatility-adjusted stats, both real estate dev stocks significantly outperform the FMCG ones. From a portfolio perspective, I would be inclined to argue that the real estate companies were better picks and contributed more.

One refrain that sceptics may be holding onto is that, as per the NAV charts, the volatility in the real estate stocks is still too high. So, let’s look at it differently. Below we compared the share price CAGR from various points in time. We see that regardless of when you invested in a 10-year period, if you invested in the real estate stocks, you did better than if you invested in the FMCG ones. From a portfolio perspective, the real estate companies are now clearly better picks.

What fundamentals and screeners miss

Where is the logic behind this? We have spent so much time previously studying valuations and financial theory. There are people much smarter than us handling a lot more money than us to allow such mistakes to happen in the market. How is it that the fundamental screeners and ratios miss the forest for the trees so badly in this specific case? Comparing the assets of the companies we see what’s missing or, rather, we don’t see what’s missing.

Even the assets show that there seems to be no undue gain for the real estate companies that would justify such superior stock performance. So we’re seeing worse sales, worse/no sales growth, more volatile margins, lower operating profit and CFO growth, and no disproportionate asset growth. Then where is the outperformance coming from? And why does GODREJPROP perform even better than DLF, when DLF has the better ratios of the two?

The answer lies in land banks: land that was purchased a long time ago and is now monetisable either through direct sales or through new project construction. And of the two companies, GODREJPROP has much larger land banks whereas DLF generally purchases new land or partners with other landowners for its projects. The reason that land banks don’t show up in fundamentals or assets as much as you would expect is that land is valued at acquisition price on the balance sheet. It only flows into the PnL once the land is either sold or units developed on the land are sold. This, unfortunately, will never show up in any fundamental screener or ratio.

Apples and oranges

We knowingly compared apples and oranges here to highlight a specific shortfall that we see in popular portfolio construction: portfolio managers routinely compare apples and oranges every day without thinking twice. However, the metrics that apply to one sector may not apply to another. And the market consists of many different sectors. Thus, the over-reliance on tools like fundamental ratios leads to sub-optimal portfolio construction.

Let’s say a portfolio manager’s mandate is to deliver the best risk-adjusted returns that he can. His method, however, should not be to only pick the “best companies” based on popular metrics. This is because such a process tests all candidates on cookie-cutter parameters that some of them simply may not be suited for.

This is not to say that FMCG companies are bad or that good fundamentals are undesirable. We merely highlight that the one-size-fits-all approach doesn’t work when it comes to building a good portfolio. This is especially true in the modern world where new types of businesses and business models are being developed every day.

As portfolio managers, we need to remember that:

• Accepted practices may and biases be inherently misleading us to compare apples to oranges.
• Certain parameters may implicitly favour certain sectors and will always screen out others, as we have just seen.
• In the end, a stock’s risk-adjusted performance has to be a large consideration of whether it should be added to a portfolio or not.
• Eventually, portfolio performance matters more, not whether the constituent ratios were the best or most predictable.