Baidu’s AI Cloud Helped Lift Its Q4 Results And Could Lead To More Valuation Gain If The Shift Continues

Summary

• Baidu announced encouraging Q4 results driven by non-marketing Baidu Core businesses, such as AI Cloud.
• Its revenue mix shift (from online marketing to non-marketing) from 2017 to 2020 indicated a successful trend of restructuring.
• Such revenue shift is expected to accelerate in the future which could drive the multiple expansion to achieve an IRR >20% in the next 3 – 5 years.

Photo credit: Tomohiro Ohsumi, Bloomberg, Getty Images 

Baidu (BIDU) announced encouraging Q4 results last week, with its share price up 14% last Friday.

Although Baidu’s Q4 revenue only had a mild growth of 5% year-on-year, it was largely driven by an encouraging 52% year-on-year growth in non-marketing revenue from Baidu Core businesses.

Baidu Core non-marketing businesses include AI Cloud, Intelligent Driving, Smart Assistant, etc. Among them, AI Cloud’s revenue grew significantly by 67% year-on-year in Q4. The strong growth in AI Cloud and other non-marketing businesses indicated that Baidu’s transformation initiatives are starting to bear fruit.

Although Baidu's autonomous driving (AD) initiatives are still far from monetization, it is clearly becoming a leader in China's AD sector based on its number of testing vehicles and testing mileage. As such, Baidu is very well poised to benefit from the sector’s future development in China, which has been strongly promoted and supported by the government.

The integration of YY's live streaming business upon its closing is expected to diversify Baidu's business further away from online marketing. Management also gave strong guidance on the Q1’21 topline, which indicated a year-on-year growth of 15% - 26%.

Consistent revenue mix shift showcases the trend of its transformation

The following chart showed Baidu’s revenue mix (online marketing revenue vs. non-marketing revenue) from 2017 to 2020.

Source: Baidu company disclosure

From the chart above, we can see a clear business restructuring trend where the contribution of the non-marketing (non-search) business more than doubled from 2017 to 2020.

Based on the trend, it is fair to assume that the non-marketing revenue mix could go as far as 40% in FY’21, especially given that the company’s initiatives in AI and AD are really starting to bear fruit.

Baidu’s valuation is driven by the non-search business

The following chart showed the leading P/S multiples of Baidu and Sogou (SOGO). Sogou is a pure-play online search company in China.

Source: Capital IQ

Although Baidu was trading at a premium to Sogou, thanks to its dominant market position in the Chinese search sector (Baidu's 76% market share vs. Sogou's 11%), we can see that Baidu has actually been priced similarly to Sogou for the majority of the time.

What then set Baidu apart from Sogou was indeed thanks to its rumors to enter the electric vehicle (EV) market last December, as we can see from the increasing premium Baidu started to trade above Sogou since then.

For the first time over the past three years, the market is finally pricing Baidu differently. And even so, Baidu is only trading at the level it once traded at three years ago.

What does that all mean?

In my very first article (Baidu Is More Than A Search Engine And How Its Current Valuation Has Not Priced That In), I used a SOTP (sum of the parts) valuation methodology to value Baidu. Although the SOTP methodology is, no doubt, the most appropriate method to value Baidu, it is usually not easy to do due to the limited disclosure of Baidu’s business segments.

Many business segments, such as AI and AD, are still at a nascent stage of commercialization and monetization. Hence, valuing these segments requires a lot of assumptions and judgment calls that one lacks or takes time and sector expertise to develop.

In my opinion, instead of doing a full-blown SOTP valuation, we can also adopt a much simpler approach based on its revenue mix.

Applying Sogou’s 2.5x P/S multiple to Baidu’s FY’20 online marketing revenue of $11 billion gives us a roughly $28 billion search valuation.

Baidu currently has a market capitalization of $115 billion and a net cash position of $14 billion (as of 2020 year-end). This implies a valuation of $73 billion for its non-search business. Given Baidu’s FY’20 non-marketing revenue of $5 billion, the implied P/S multiple for the non-search business is roughly 14x.

Source: Company disclosure, Capital IQ

It should be noted that the $73 billion non-search valuation also included Baidu’s stake in iQIYI and Trip.com. We should take them out in theory, but for simplicity, I did not since we are only trying to find out the implied P/S multiple as a reference.

In the article “A Look Into Baidu’s Long Term Return Potential”, we established that investing in Baidu (even at its current historical high price level) could still generate a return of 20% - 25% per year for the next 3 – 5 years, provided that it could successfully transform itself into a company that focuses on AI / AD / EV, which will then be followed by a multiple expansion from currently 6x P/S multiple to 12.7x (a level which AI / AD / EV companies are currently trading at).

As we can see, the 12.7x multiple is very much in line with the implied multiple of 14x on Baidu’s non-search business. Therefore, we can argue that if Baidu’s revenue mix can continue to shift to the non-search business, a multiple expansion to 12.7x is definitely within reach.

Hence, simply by tracking Baidu’s revenue mix, we can roughly know how much more upside there still is and how the investment thesis is being played out.

Now what?

With its non-search business contributing 32% of its total revenue in FY’20, there clearly is still upside and, frankly, quite a long way to go, although non-search contribution has already been doubled since FY’17.

In my opinion, the uptick in the non-search revenue contribution will not be linear. It will only accelerate further in the future since many initial investments are needed prior to full commercialization and monetization, as in the case of AI, AD, and EV. A long term investment horizon is definitely needed to realize the upside fully.

It took the market more than three years to realize that Baidu, with more than 30% revenue from its non-search business, is more than just a search engine. With so much positive momentum going on in Baidu recently, I am hoping it would take much less time for the market to realize that Baidu is laying so much groundwork now to become a deep tech company in the future.