A few years back, we wrote about Our Worst Investments in our Q2 FY21 quarterly letter. The letter talked about all the investments where we lost more than 15% of our invested capital. Each of those investments hurt our returns and dragged down the overall portfolio performance. In these cases, it was our buying decision that turned out to be wrong.
In this letter, we talk about another kind of decision that has had an even larger impact on the portfolio returns – the selling decision. Since our fund’s inception, 6 out of our 25 exited investments have delivered returns of 5-15x post our exit. If we hadn’t exited these investments, our portfolio returns would have been substantially higher. We discuss below these 6 investments – the investment/selling rationale, what we missed, and learnings, if any.
Safari Industries
We invested in Safari, the 3rd largest luggage player in India, in 2016. We believed that the organized luggage industry would benefit from the growth of air travel, and from the unorganized to organized shift. Safari was led by Sudhir Jatia who had a phenomenal track record in the luggage industry. He had previously sold Aristocrat to VIP and then led VIP to greater heights as its Managing Director. After the acquisition by Sudhir Jatia in 2011, Safari had turned around and was consistently gaining market share against VIP and Samsonite. Our thesis was that Safari would continue to grow rapidly with expanding margins, and was likely to be the biggest beneficiary of the industry trends.
Our investment thesis panned out well. Safari reported strong revenue growth and margin expansion post our investment. The Safari stock performed even better, with valuation multiples expanding due to rising investor interest. Safari had been a large position, and its success led to the overall fund doing well.
In 2018-2019, we exited our Safari investment for a ~4x return. The main factor that drove the sell decision was the valuations, which had crossed 80x TTM1 P/E. We believed that the future growth and margin expansion might be lower as Safari had already reached a reasonable scale, and incremental market share gains might not be easy. This perspective turned out to be wrong. Barring the Covid dip, Safari’s business has done phenomenally well. Competitors like VIP have struggled due to management issues, while Safari’s strong execution has led to substantial market share gains. Safari is now more profitable than VIP. The strong business performance resulted in the Safari stock rising 5x post our exit.
We clearly underestimated Safari’s growth potential despite holding Sudhir Jatia’s leadership capabilities in high regard.
Tata Elxsi
We invested in Tata Elxsi in 2019. At that time, Tata Elxsi was facing business challenges as its largest and most profitable customer, Jaguar Land Rover, was struggling and reducing its R&D outsourcing budget. As revenues and margins declined, the Tata Elxsi stock fell by over 50% from its previous highs. We had been following Tata Elxsi for several years and liked the ER&D Services industry. In our view, Tata Elxsi’s challenges were temporary as the non-JLR business was still growing at a healthy rate. With valuations declining to 18x TTM P/E, we found the risk-reward quite attractive.
Tata Elxsi’s troubles turned out to be short-lived. Growth in other business segments compensated for the decline due to JLR, and Tata Elxsi’s margins expanded materially. As the business momentum picked up, the stock recovered and begun trading at valuations north of 40x P/E. We found the valuations excessive as we believed the 29% EBITDA margins may not be sustainable and growth may not exceed 15%. Our anchoring bias of having paid only an 18x P/E might have also played a role here. We exited the investment in early 2021 at an attractive ~3x return in less than 18 months.
To our surprise, Tata Elxsi over the next 2 years, reported >30% growth in revenues and margins improved further. The stock went up over 5x from our exit price as the valuations crossed 100x P/E! Investors who were earlier worried about the JLR issues and unwilling to buy at sub-20x P/E were now extrapolating strong growth into the future and willing to pay any price. Although growth has now slowed down, the stock still trades at >60x P/E (~4x of our exit price).
While selling stocks like Safari and Tata Elxsi may seem like a massive mistake given their post-exit stock performance, we think such exceptions are normal for any investment strategy that believes in selling stocks for valuation reasons. Every now and then, an extraordinary entrepreneur like Sudhir Jatia may make concerns around valuation seem trivial, but that’s okay. In most other cases, selling when valuations become excessive will be the prudent thing to do, as often even good businesses will struggle to justify such lofty valuations. This was seen in the performance of some of our other exits due to rich valuations, where subsequent stock price performance has been disappointing.
Saregama India
We invested in Saregama in 2018. Our investment thesis was that the rise of OTT music streaming apps like Spotify, Saavn, YouTube, Apple Music, etc will result in a significant increase in legal consumption of music in India. We expected Saregama to benefit from this trend as it owned the IP for a large library of music in Hindi and other Indian languages. Saregama also had a large share of its revenues coming from Carvaan, a retro portable pre-loaded music player, which was selling over 250,000 units a quarter. Our investment thesis did not ascribe much value to the Carvaan business (despite it being more than 50% of revenues at the time of our investment).
Post our investment, both Carvaan and the Music Licensing business showed good growth for a few quarters. However, Carvaan soon started to stagnate despite new product launches and increased marketing spends. This did not worry us much because we believed that Saregama would still be able to operate Carvaan profitably at a smaller scale. However, the Management / Promoters seemed to view the slowdown as a minor blip. Instead of cutting costs, Saregama ramped up its sales and marketing spends, even resorting to expensive media such as TV. Carvaan from being a profitable business was now expected to lose money if sales did not pick up. This concerned us. Saregama already had another business – print publishing, which had been losing large amounts of money for years. It seemed that Carvaan was set to suffer the same fate.
The increasing likelihood that the good cash flows from the Music Licensing business would be used to subsidize other bad segments forced us to rethink our investment. In 2019, we exited Saregama at a loss of over 30%2.
While the loss hurt us, what happened subsequently has been even more painful. As per our expectations, Carvaan sales continued to decline despite aggressive marketing. However, the Music Licensing business did exceedingly well, growing at a 20%+ CAGR with improved profitability. After seeing the continued weakness in Carvaan sales, the management promptly reduced its marketing spends on Carvaan and is now operating it on a breakeven basis. Saregama also decided to demerge its non-core assets including the loss-making print business. Saregama profits over the last 4 years have grown 4x+ and the stock is up ~10x from our exit price.
In this case, we clearly made a mistake by selling due to a single adverse event even though our core investment thesis was intact. We didn’t account for the positives, nor did we consider management’s ability to course correct. We are now far more patient and unlikely to have a knee-jerk reaction to bad news and decisions. In most cases, management can reverse these decisions if things do not go as expected. We reinvested back into Saregama in 2022 at ~6x of our exit price!
Tata Investment Corporation
We invested twice in Tata Investment Corporation (TIC) – once in 2016 and again in 2018. The TIC investment followed our HoldCo Investment Framework – invest in holding companies that have a high quality underlying portfolio, a good dividend yield and trade at higher than average HoldCo discount.
In India, most holding companies trade at a HoldCo discount of 50-80% of the underlying holdings’ value. Even TIC has historically traded at a discount of 55-60%. TIC holds stakes in some of the best Tata group companies and pays a sizable dividend. In both cases, we invested in TIC when it was trading at a discount of >60% and exited when the discount narrowed substantially. Through this process, we benefitted from an appreciation in the underlying holding value, a high dividend and a narrowing of the discount. We generated high returns in both instances.
However, we could have made far higher returns had we not sold TIC in 2021. Since then, TIC has benefitted from a strong rally in its underlying holdings. Tata group stocks like Titan, Trent, Tata Motors, Indian Hotels, etc have contributed to a significant increase in the value of TIC’s holdings. But one surprising development is that TIC today trades at a ~15% premium to the value of its underlying holdings (a few weeks back the premium was ~50%). This seems like a bizarre situation. There is little reason for investors to pay a premium to the underlying holding value when they can easily buy the underlying stocks and replicate most of TIC’s holdings. Some of the euphoria may stem from TIC’s ownership of a stake in Tata Sons which was rumoured to be going public. However, TIC’s stake in Tata Sons is minuscule and even assigning a large value to Tata Sons wouldn’t justify the 15% premium.
While TIC has seen a ~5x increase post our exit, we wouldn’t consider it as a miss. Staying invested in TIC would have required us to abandon our HoldCo investment framework and to hope for the stock to rise to irrational levels far disconnected from its fundamentals.
Astra Microwave
We invested in Astra Microwave in 2016. Our thesis was centred around Astra benefitting from an increase in indigenous defence spending and the strength of its radar technology capabilities. Unfortunately, Astra didn’t show any immediate signs of improvement in defence ordering that could benefit it. Our subsequent channel checks suggested that any material increase in domestic defence spending still faces several barriers, with the primary one being long approval cycles for new products. As our conviction was weak and without any material improvement in Astra’s business performance, we exited Astra in 2017 at almost breakeven.
Since our exit, Astra’s business has largely struggled– profits stagnated, and the balance sheet deteriorated. The stock also didn’t appreciate for many years. However, in the last 3 years, Astra’s stock price has seen a 6x increase. This has been largely driven by the euphoria surrounding the defence opportunity which has benefitted all defence stocks. In recent quarters, Astra’s business has also shown some signs of improvement.
While the Astra stock is today at 6x of our exit price, it is clear we wouldn’t have been able to remain invested when the business struggled for such a long time. With valuations of 75x P/E, Astra’s current rise seems to be largely driven by the bullish defence narrative. It remains to be seen if Astra is able to exploit the defence opportunity meaningfully.
Ion Exchange
We invested in Ion Exchange in 2016. Our investment thesis was based on the sustained growth of the highly profitable water chemicals segment. The Company had significant market share in the resins business and made supernormal returns. Although this segment contributed less than 50% of overall revenues, it accounted for more than 70% of the profits of the company. We were less excited about the capital-intensive engineering division (EPC-led), characterized by lower margins and primarily serving government and quasi-government entities through L1 tender-based contracts.
We exited our position in the stock at breakeven following an unfavorable SEBI ruling. The ruling demanded the company to reimburse funds in a 13-year-old illegal collective investment scheme case. The amount seemed excessive and would have wiped out many years of accumulated profits. We sold the stock without assessing the likelihood of SEBI’s ability to enforce the decision. While the amount was substantial in relation to its profits, it was not very material in relation to its valuation. Here again, like Saregama, we let a non-core issue affect our investment thesis on the core business. The stock has since risen by more than 15x. The chemical business performed better than our expectation and continues to be the main contributor to the company’s value. The company has challenged the SEBI ruling, and its resolution is still pending.
Like with Saregama, the key learning here has been to understand adverse events in more detail before reacting. In the last quarter, we wrote a note on the folly of selling on bad news. The note incorporates our learnings from past mistakes like Saregama and Ion Exchange.
Analyzing past investment decisions is a tricky task. It is often unclear whether a bad outcome was due to a bad process or simply bad luck. But this challenge alone shouldn’t stop us from dispassionately studying our past actions. While in many cases there may be nothing to learn, occasionally, we may be able to rectify flaws in our process that improve future results.
_________________________________________________________________________________________________________
1 Trailing Twelve Months
2 We had also written about Saregama in our Q2 FY21 letter
