Description

Summary: Prosus trades at a 43% discount to NAV. 75% of net assets is Tencent, which is a quality, growing business that looks itself undervalued. Importantly, a catalyst to narrow the discount exists in the from an open-ended buyback program, which management seems very committed to. Management seems to target a reduction of the discount to under 35% of NAV which I expect them to achieve within two years. Even if NAV ex buyback remains unchanged during that period, NAV per share should increase by 5-10% because of the buyback and in combination with the discount reduction would result in an IRR 10-15% p.a. incremental to what the assets generate over that timeframe.  

History 

I assume most members will be familiar with the name, but for those who are not below a brief history for context. 

About 20 years ago a listed family media business in South Africa called Naspers hired a telecom executive, Koos Bekker. He agreed with the family to forgo a salary and work exclusively for a share of the value he would create investing. Subsequently in 2001, Koos bought a 46% stake in Tencent for a few million dollars which turned out to be one of the best investments of all time and made Koos Bekker a billionaire and ultimately Chairman of Prosus. Naspers went on to sell a portion of the stake over time and were diluted through a few events as well bringing their stake to slightly below 30%. 

Naspers continued to be listed in South Africa and soon became a very large portion of the index. North of 80% at some point I think and as a result performed poorly to the point that Naspers excluding its Tencent stake was trading at a negative value. 

In 2019, Naspers management decided to try to rectify this problem and to list the Tencent stake and other investments separately in a new vehicle called Prosus in Amsterdam. To minimize taxes for South African shareholders a cross shareholding was established. The discount to fair value moderated slightly but continued to much higher than what management aspired to.

In 2021, Prosus implemented a share swap between Prosus and Naspers to tighten the discount, but again the discount to NAV persisted. 

In June 2022, Prosus announced a buyback program funded by a disposal of Tencent shares. The timeframe and the amount of the buyback program is open ended. The discount reduced a little as a result from mid fifty percent to 43% today. 

Holding Discount 

Holding discounts often exist for good reasons and Prosus is no exception. However, the current 43% level seems hard to justify. Let’s look at some of the potential explanations in detail: 

Lack of a Catalyst 

Holding company discounts are frequently very persistent because there is no catalyst to change the status quo. In Prosus’ case however I think there is a catalyst. As discussed above, Prosus announced an open-ended share buyback. To date Prosus has repurchased about $3.9 Bn (about $600 MM of which were Naspers’ shares which has the same economic effect as buying Prosus shares.). Management talks about the buyback program extensively and seems quite serious to do “whatever it takes” to bring the discount to below 35% to NAV (below the first three slides from their earnings announcement deck). Note the illustrative sizes they are throwing around for the share buyback. $10-30 Bn seems sizable for a business with a market value of EUR 73 Bn. Also note that narrowing the holding company discount would trigger an incentive payment for both CEO and CFO of $2-3 MM each vs. total comp of $4-6 MM ex stock options. 

Capital Gains Tax

Prosus’ key asset are Tencent shares. I don’t know the exact tax basis, but it’s likely effectively zero given the history. According to management, the Tencent shares are held through a Dutch subsidiary and are not subject to capital gains tax. I am not a tax expert, but there are two reasons that give me comfort that this is correct. First, Prosus sold shares in Tencent in the past and did not pay taxes. Second, there seems to be a provision in the Dutch tax code that exempts any holing above 5% form capital gains tax. Note that once Prosus falls below 5% of ownership of Tencent capital gains tax would apply at the ordinary corporate income tax rate of currently 25.8%. Dividends seem to be likewise exempt from taxation provided the holding is above 5% of shares outstanding. 

In summary, about 18% of Prosus stake in Tencent could be subject to taxation at 25.8% which would justify a c.4% discount. However, this is based on two assumptions which are probably overly conservative. First, Tencent has its own share buyback program meaning that the proportion of Prosus’ stake subject to taxation will ultimately be lower than what the current percentages suggest. Second, there are several ways to reduce the capital gains tax bill through deductions and the like. How much is difficult to quantify but assuming Prosus would have to pay the full amount is probably punitive. 

Holding Costs 

Estimating the holding costs is not straightforward. While the majority of Prosus value is in equity accounted Tencent, many of its unlisted investment are consolidated. As the result the group has 26,000 employees. The listed TopCo is required to file unconsolidated accounts but has no employees. The majority of holding costs are likely management compensation which runs at $84 MM in FY 2022 and $130 MM in FY 2021. I am using “corporate” SG&A from the segment disclosure as a proxy which is $167 MM in FY 2022 and $110 MM in FY 2021. Capitalizing that at a somewhat arbitrary 25x equals $4.2 Bn.

Capital Allocation 

Historic returns have been very good, but almost entirely driven by Tencent which seems to have been a lucky punch. So, what about the other investments? Prosus has an IRR target of 20% and claims that up until FY 2021 they were at 22% since 2008 excluding Tencent. In FY 2022 that IRR figure came down to 13%. However, the fundamental problem with these numbers is that there is no way to verify them independently and many of the investments are private benefitting from inflated pre-IPO rounds and Prosus’ subjective marks. The only real alternative is to look at how individual investments have done which is only possible once they IPO. One of their biggest investments has been Delivery Hero and based on the public disclosures Prosus average costs seems to be slightly north of EUR 60 vs. a current share price of c. EUR38 which for a multiyear investment with venture capital type risk seems very poor. There are smaller counter examples such as Remitly where Prosus average share price seems to be in the range of $5 vs. a current share price $9.81 which looks respectable though not fantastic either. Other investments such as OLX, Swiggy and PayU (ex Bill Desk acquisition) seem like they should have been successful, but without any way of quantifying. 

Management hinted in the last earnings call that going forward they will prioritize the share buyback and ‘value crystallization’ over additional investments in the venture portfolio, but it is important to recognize that the venture portfolio is the ‘raison d'être’ for management, so I would not expect investments into this area to stop entirely. 

In summary, the recent focus on share buybacks is a very good sign, but while management’s track record did have some real successes the overall record seems mixed and not easy to assess. On the other hand, while I am not sure that management has created a lot of value, they don’t seem have destroyed a lot of value either. Therefore, I haircut the existing value of the private portfolio (see below) and I assume that future spending is neutral (which I recognize is an assumption you can have a real debate about). 

Complexity

The legal structure and cross holding between Prosus and Naspers are complicated and messy. That said, the result is straightforward. There are about 1.4 Bn shares entitled to a pool of assets which mostly consists of Tencent shares and some unlisted investments. While I admit that some discount is probably warranted to compensate for this complexity, I think the structure is far easier to understand than the Bollore complex for example. So, what is the right number? I would feel comfortable with a 20% discount to account for the questions around the capital allocation record and the complexity, but that is admittedly very subjective.

Unwanted Diversification, Illiquidity Discount  

Often when you buy a holding company you are forced to buy assets you would not buy on a standalone basis. In Prosus’ case however one asset, Tencent, accounts for 84% of the NAV if you mark the unlisted assets at 50 cents on the dollar as I do. Also note that unlike many holding companies Prosus’ assets are mostly listed and do not warrant a discount to compensate for their illiquidity. 

Asset Marks

As discussed, the majority of NAV are listed assets with decent trading liquidity. However, a little over a quarter of NAV are unlisted assets mainly VC type investments. Prosus’ says they are using street estimates where possible, but the valuation is not transparent. Further, since Sep-21 valuation has decreased by only about 20% while public market valuations of high-growth business have fallen much farther. So, I think a haircut is warranted. How much? Again subjective, but 50% seems fair to me which would bring the decrease more in line with the decline in high growth stocks. 

Valuation 

Helpfully Prosus publishes daily NAV calculations on their website, which is a good starting point if you want to look into this idea. Below the current NAV as well as the development over the last few quarterns. 

Going forward there are three sources of potential NAV per share accretion: 

First, through ‘organic’ NAV growth. This comes mostly down to the share price development of Tencent. I think Tencent is cheap, but it would require a whole different write-up to cover that topic properly. So, to make life easier let’s assume no organic NAV growth and that Tencent is fairly valued.

Second, the narrowing of the discount to NAV. Why do I think that the discount will narrow? Because management can effectively force the issue if they size the buyback large enough which they seem prepared to do (have a look at the sizes they use to illustrate the buyback in their earnings presentation). During the last earnings call management pegged the ‘acceptable’ discount to below 35% to NAV. Unlike many holding companies most of Prosus assets are liquid, so management also has the means to follow through. I believe they will achieve their goal.

Third, NAV per share will increase because of the share buyback. Prosus’ shares currently trade at 43% discount to NAV, and I believe management will buyback shares until the discount hits 35%. Prosus has historically done two $5 Bn buyback in 2021 and 2020. Both did not narrow the discount materially. So, I think it will take $15 Bn+ to achieve their goal spread over two years (currently the buyback programs is running at around $11 Bn annualised). It is unknowable how much Prosus will pay for the shares it will acquire, but if you assume that the share price gradually appreciates as they work through the buyback EUR 59.54 price per share on average seems reasonable which implies a 38% discount to NAV. As a result, NAV per share would increase from currently EUR 97 to EUR 106. A 35% discount to that NAV number - which is management’s target - implies a EUR 69 share price or a 11% IRR over two years. Note that at that point if you marked the unlisted assets at fifty cents on the dollar and adjusted for holding costs and potential tax leakage etc., NAV would be around EUR 84 implying a 20% discount vs the then implied share price of EUR 68, which seems about right. 

Risks

Tencent: All risks that apply to Tencent (VIE structure, China related risks etc) also apply to Prosus 

Capital allocation

Change in taxation   

Pace of share buybacks slows, or it takes longer to narrow the discount (though this would likely be partly offset by a larger buyback though would still result in lower IRRs)

Why not Naspers? 

The discount to NAV is larger for Naspers, so why not buy that? I think you can make valid arguments for both Naspers and Prosus. The reason I prefer Prosus is that management has shown to treat Prosus slightly more favourably (have a look at the recent exchange offer for example), Naspers has another management layer and some South Africa country specific risks as well and lastly the tax implications of owning Naspers shares seems more complicated esp. if the there is a big change in the corporate structure. Management is more exposed to Naspers share price via stock options, but on market insider purchases seem to have favoured Prosus. 

Disclaimer 

We own this (surprise!). 

We have tried to take care in preparing this write-up, but you know, it could be all wrong.

We may sell or change our mind and we won’t tell you about it.

Catalyst

Share buyback narrowing discount to NAV while growing NAV per share