Description

Better than holding cash and potentially more… 

We are revisiting SoftBank Group, a company we wrote up six years ago (February 2014).  At the time, we viewed SoftBank as an inexpensive way to create the soon to be public Alibaba Group.  Since that time, the stock is up roughly 30%, while its market cap has grown from roughly $90 billion to $104 billion and the company’s share count has decreased by over 13%.  The underperformance of SoftBank shares can be attributed to two factors: (1) the market’s preference to own pure-play operating companies, rather than sum-of-the-parts stories; and (2) a large sample size of poor capital allocation decisions at the Vision Fund, the most spectacular failure being WeWork.

In our experience, sum-of-the-parts stories are tough ways to outperform the market.  Typically, they are trading vehicles for times when spreads get too wide and there is a short-term catalyst at hand.  

We began working on this write-up at the time of Elliot’s announced involvement in SoftBank, as we believed this could be a value-unlocking catalyst.  We believe Elliot may have some success pushing SoftBank to improve its capital allocation (i.e. less Vision Fund 2 and more aggressive share buybacks, something SoftBank is no stranger to pursuing.)  

We also thought we were getting an inexpensive call option on the possibility of the TMUS/S deal going through.  While SoftBank shares have already rallied materially on both developments, we believe Elliot’s involvement combined with a successful TMUS/S transaction meaningfully improves the chances of SoftBank pursuing meaningful share repurchases.  

With SoftBank trading at just 56% of mark-to-market net asset value (using conservative marks for non-public holdings), we believe the shares are set up to deliver attractive risk adjusted returns. 

To provide some long-term perspective, in February 2014, we valued SoftBank’s 33% stake in pre-public Alibaba at $66 billion, 60% of the market cap of the SoftBank at the time.  Today, SoftBank's since reduced 25% stake in BABA is valued in the public market at $147 billion or 142% of Softbank’s market cap.  

As mentioned above, on a mark-to-market basis, SoftBank trades at 56% of its net asset value of $184 billion or $45 per ADS.  Valuing BABA at a reasonable 12-month forward estimate of $250 per share suggests that SFTBY trades at less than 50% of our $50 per share estimate of net asset value.  

Our 12-month target of $32 per SFTBY share assumes the following:

1. Alibaba Group. A target price of $250 for BABA and a 30% hold-co/minority interest discount.

2. SoftBank Corp. Current share price of Y$1502 for SoftBank Corp. (9434 JP) and a 30% hold-co/minority interest discount.

3. Sprint. Current share price of $99.50 for TMUS; 3.446 billion Sprint shares owned by SoftBank; a revised S/TMUS exchange ratio of 11.31:1 for SoftBank’s Sprint holdings; and a 30% hold-co/minority interest discount.

4. ARM Holdings. A 70% valuation discount to the 2016 purchase price of $32 billion for ARM holdings; 75% of the position is held at SoftBank parent. 

5. Vision Fund.  A $10 billion valuation for Vision Fund vs. SoftBank’s current carrying value of $28 billion

6. No other value

Our base case scenario assumptions (30% discount on all publicly-traded assets) imply a 28% discount to total target gross asset value ($188 billion vs. $260 billion) and a 35% discount on a net (i.e. after net debt) asset value basis - $134 billion vs. $206 billion.

SoftBank: Sum-of-the-Parts Analysis, BABA at Market

SoftBank: Sum-of-the-Parts Analysis, BABA at $250

Why Now?

Since the stock’s bottom on October 28th, when the shares traded at a 53% discount to mark-to-market net asset value, we believe several positive changes have developed, particularly in recent weeks.  We believe we are in the early stages of a shifting in the SoftBank narrative for the better. The positive developments include the following:  

1. Masa Son is no longer writing blank checks for Vision Fund portfolio companies.  Vision Fund is not a zero and we believe is likely worth at least $10 billion to SoftBank’s equity. 

2. TMUS – S will soon close and SoftBank’s stake in Sprint has recently re-rated by over $15 billion (more than $3.50 per SFTBY share) or 14% of SFTBY’s current share price.

3. Elliot Capital Management announced that it has made a material investment in SoftBank, estimated at close to $3 billion.  Press reports have suggested Elliot wants SoftBank to take advantage of the stock’s wide discount to net asset value by purchasing $10 to $20 billion of stock.      

SoftBank’s Key Assets and Valuation:

SoftBank today has a $157 billion enterprise value and a $103.5 billion market cap. Masa Son owns ~27% of the company, currently worth ~$28 billion.

• Alibaba stake (BABA, 25% stake): BABA is worth ~$147bn to SoftBank at market.  SoftBank has sold some of its shares over the past few years, recently settling a forward sale agreement in June 2019 (from 2016) for ~73m shares worth >$12bn.  Masa San appears reluctant to pursue further near-term monetizations of SoftBank’s BABA holdings given his bullish view of the asset. That would be fine with us as we are constructive on the long-term growth outlook for the company.  We value BABA at 20-25x calendar 2021E EBIT, which yields a forward valuation above $250 per share. Obviously the COVID-19 virus will continue to severely impact China’s growth in the short run and BABA’s business is being impacted negatively.  Our view is that by the end of 2020, the market will be valuing BABA based on 2021 estimates, which will likely reflect a meaningful recovery in the business and a normalized operating environment.  

• SoftBank Corp (9434 JP, 66.5% stake): SoftBank Corp. represents the SoftBank parent’s holdings in Japanese domestic telecom and Internet assets.  After BABA, SoftBank Corp has been the best investment that SoftBank has ever made, as measured by shareholder value creation for SoftBank.  The stock pays a 5.6% dividend (at current prices) or about $2.4 billion (based on current US$/Yen exchange rates), which SoftBank uses as one of its primary sources of funding.  

• Sprint/TMUS (TMUS, 24% stake): Based on a revised TMUS:S exchange ratio of 1 to 11.31, TMUS closing share price of $99.50, and SFTBY’s 3.446 billion shares of Sprint, SFTBY’s stake in the new TMUS can be valued at $30 billion.  This implies a price of $8.80 per Sprint share. The public shareholders continue to receive 1 share of TMUS for every 9.75 shares of S. However, SoftBank will now receive 304.7 million shares of TMUS, down from 353.4 million shares under the prior terms.  Notably, there is a path to recover those 48.8 million shares; if TMUS reaches a value of $150 per share by 2025 or $160 per share by 2026, the shares are returned. This contingent value right is worth $7.3 billion (i.e. 48.8 million shares x $150 per TMUS share); it is quite significant.  There is a 4-year lock-up on SoftBank’s 24% pro-forma holding in TMUS. We expect SoftBank to borrow against its holdings in TMUS going forward, perhaps raising as much as $5 to $7.5 billion against $30 billion in mark-to-market value. We believe our Sprint valuation to be potentially conservative for several reasons: (1) We assume no further appreciation in TMUS share price; (2) We apply a 35% valuation discount to the holdings, despite the shares having a tax basis north of $5.25 per share and our expectation that SoftBank will likely monetize the holdings over time; and (3) We assign zero value to the CVR, which could easily be worth $7.3 billion in 5 years.     

• ARM Limited (75% stake, private): ARM is a microprocessor company that largely sells its IP to semiconductor companies.  SoftBank bought ARM in 2016 for ~$32bn. Since that time, it has sold ~25% of that stake to the Vision Fund.  ARM is a high margin licensing business undergoing a period of heavy investment spending which has yet to translate into meaningful revenue growth.  The company has signed 600 licenses since Q1 2015, many of which have yet to materially contribute to revenues. Moreover, the company has recently changed its licensing policy on its most popular processors to forego some upfront licensing fees in exchange for royalty revenues.  We value SoftBank’s 75% stake in ARM Holdings at 5x projected C2020 revenue of $2 billion or $7.5 billion, a 70% haircut to the value that SoftBank paid 4 years ago. For more on Arm Limited click here.  

• Vision Fund: SoftBank has a capital commitment of $28bn to the Vision Fund, with $23bn of equity capital already invested. It is also the GP, earning management fees and a 20% performance fee. What SFTBY’s equity in the Fund is worth is subject to much debate, but we will go through this in more detail in the appendix that follows.  At a high level, we assume a 58% discount to the private portfolio and use FMV of all public investments as of yesterday.

• Fortress Investment Group (100% stake): SoftBank bought all of Fortress in Dec 2017 for $3.3bn. We value that business at >50% discount to the purchase price to be conservative.

Vision Fund I: 

Recent presentation is found here.

In Oct 2016, SoftBank announced it was launching a $100bn technology fund (ended up with $97bn total). In addition, SoftBank had previously launched a Delta fund for $6bn. The company usually combines the two funds. The terms of the Vision Fund were unique with a certain portion of the capital commitment coming from a “preferred” piece and the remainder from “equity”. Key terms as follows:

• Every 3rd party investor in the Vision Fund invested at a 1.57x preferred to equity ratio. This served two purposes, 1) to provide 3rd party investors with an LP interest higher in the capital structure, with current income, lower potential upside, and more safety; and 2) to provide a levered return to SoftBank if the Fund was successful.

• The preferred pays a 7% current coupon. In a waterfall, holders of the preferred receive their principal back before the equity holders receive any of their principal back. But the preferred return is capped at a 7% annual return on principal.

• The equity piece is subject to an 8% hurdle rate and pays out 20% to SoftBank as the GP for any profits in excess of that. SoftBank, as the largest equity holder (57% including employees), and the owner of the carried interest, is materially levered to any success of the fund beyond the initial $39bn of preferred (plus dividends) value. 

• There is leverage upon leverage for SoftBank in the Vision Fund (equity piece + carried interest). But SoftBank is not liable for anything beyond that which they contribute as part of their committed capital ($28bn).

• In addition, SoftBank offered key employees the ability to invest in the fund by taking out debt backed by SoftBank. Much of the $5bn was from Masa Son (media speculates ~$3.5bn) with the rest effectively being used a form of compensation for other employees. It is unclear if SoftBank is on the hook for that debt if the fund completely blows up.  It is believed that Masa has used some of his $28 billion in fair market value of SoftBank stock as collateral for his investment in the Vision Fund (margin loan).

• The term of the fund is 12 years, with a two-year extension. Management fees range from 0.7-1.3% with a blended ~1%.

• There is a $4.1bn, 3 yr revolving financing facility within the fund that permits the fund to put incremental leverage on select positions (typically those that have gone public) with a 33% LTV.

As of 12/31/20, the fund was more than 80% invested across 91 different investments. The Fund plans to stop investing in new names after it reaches 85% of deployed + obligated capital, in order to leave room for follow-on investments. Some of the investments and the allocation by sector are below:

Vision Fund I Waterfall Discussion and Equity Value to SoftBank:

As of February 12th, SoftBank had a carrying value of $94 billion for the Vision Fund.  

The company has realized $8.6 billion on its holdings in NVDA and Flipcart.  

In addition, the company has $14.2 billion in public market value for its portfolio companies that have IPO’d.  On balance the portfolio is performing okay, with the total return on investment at 1.5x invested capital. If we net out the realized gains and the public holdings (in aggregate $14.2 billion) from SoftBank’s estimated $94 billion carrying value for the fund, we arrive at an estimate of $72 billion for the carrying value of the private assets and uninvested capital.  

SoftBank represents an estimated 48% of the equity funding for the Vision Fund and 28% of the invested capital.  The preferred component is estimated at $32.5 billion currently. With the inputs of SoftBank’s $94 billion in carrying value, the preferred stock and the value of the public holding and realized gains, we can estimate how much the value of the private holdings in the Fund would have to fall in order for SoftBank’s equity in the fund to be worth $10 billion.  

Based on our rough math, we believe the value of the private assets would have to decline by close to 60%.  While there are some assets that have taken enormous hits (i.e WeWork among others), there are other assets (DiDi, Grab, Paytm, OLA, etc.) that show some promise.  

Risks

Poor capital allocation

Dependence on Alibaba

Exposure to China

Catalyst

Share repurchases

More disciplined capital allocation