By Paul Bennetts via http://paulbennetts.co/
The venture capital market is probably the most inefficient of all mainstream financial markets. Valuation in venture has to rely more on a function of comparable rather than absolute valuations. Whereas, most common stocks in the public market, have an underlying business that is growing at a predictable rate tied to the overall economy and hopefully throwing off a steady stream of free cash flow. A common stock, in this case, can be valued like a bond by applying a targeted rate of return to a set of forecast cash flows.
In venture that’s not so easy. A technology business for example, may be growing fast but unpredictably, exist in a rapidly changing market or be an increasing returns business. A venture backed business will likely have a high probability of low predictability. This makes it really hard to develop accurate financial models to value its future cash flows.
A venture capitalist will instead assess an opportunity and valuation based on an opportunity set. The more deals a VC sees the easier it is for them to efficiently price a valuation based on how it fits in the universe of opportunities. Essentially, a startup needs to show its creating a more defensible business with less capital at a lower asking valuation than the next guy.
This comparable valuation method also works across time and cycles. A VC can compare today’s break out companies (unicorns) to the traction and valuations of their past peers. They could have watched as a one of these past companies entered and exited the previous bubble. In the late nineties a bubble did form resulting in the dot com crash. History, devoid of hindsight bias, can provide context for the feeling of floating in a bubble. As a result, a well-informed VC may choose to step completely out of the market (or parts of it) for a period of time.
One such past example is Yahoo. Its annual 10K SEC filings are available for review for each year following its IPO (1996, 97, 98, 99, 00, 01, 02, 03). Each of these documents can be hundreds of pages long but well worth unpacking.
The embryo of Yahoo formed in late 1993 when two Stanford students, David Filo and Jerry Yang, downloaded Mosaic the first consumer friendly browser to surf the internet. After discovering many websites, Filo and Yang created a web page listing their favorite sites.
In April 1994, their web page had hundreds of websites and was receiving approx. 1,000 page views per week. At the time, there was limited tools to discover content online. Yahoo’s product — solving discovery via human curation — fit well with the market need.
By September 1994, the team had grown its web page to approx. 2,000 links. With no paid marketing and product-market fit, Yahoo began getting approx. 50,000 hits per day.
This was obviously incredible traction but the final piece came the following month. Marc Andreessen launched an easier-to-use version of the Mosaic browser, called Netscape Navigator. As part of this, Netscape added a button called ‘Directory’ which linked to Yahoo. By January 1995, Yahoo was receiving 1m visits per day and its guide for the web now had approx. 10,000 links.
Yahoo was ready for external funding. Yahoo proposed to become an advertising supported guide for the web.
It was receiving >30m hits per month. The next year, Yahoo would be deriving substantially all its revenue from advertising. Its standard advertising rates would range from $20-$60 per thousand impressions (CPM). Today, banner ads achieve rates of $1 to $6 per CPM. Rates are in part a function of supply. In 1995, there wasn’t much online inventory.
At the time of the round though, Yahoo was pre-revenue. Sequoia may have estimated that Yahoo could produce approx. 30k advertising units per month that could be monetised for up to $5–10m per year. But Yahoo would likely need a significant BDM effort to deliver this potential.
Sequoia negotiated a deal in April 1995 to invest $1m at a $3m pre-money valuation. Sequoia gave Yahoo 24 hours to accept the proposed deal structure. Yahoo accepted.
In the year of Sequoia’s investment (1995), Yahoo generated $1.4m in advertising revenue. Incredibly, Yahoo in the following year (1996), generated $19.1m in revenue.
How does it compare to today’s valuation environment? Was this round underpriced?
Hindsight would suggest so. But the market opportunity was smaller then. The entire internet population was approx. 44.8m in July-1995. Yahoo had curated the web with only 10,000 links a few months earlier.
In my mind, this is somewhat analogistic to the potential of blockchain startups. Using wallets as a proxy for population, Blockchain.info and Coinbase have approx. 3m and 2.6m wallets respectively.
It is hard to visualise what may be ahead for the blockchain or the startups tackling it. A $1m investment in a blockchain startup could buy 25% of Yahoo or a liquidation preference.
After receiving external funding, Yahoo expanded beyond advertising by signing a deal with Reuters in August 1995 to publish ten stories per day on Yahoo. This signalled that Yahoo’s business could expand beyond web links. At the end of 1995, Reuters and Softbank together invested $5m in Yahoo at a $40m valuation (just 8 months following the Sequoia investment).
Yahoo had 6 full time employees at the time. And it was receiving 6m daily page views. Yahoo had generated $1.4m in revenue since inception and spent $738k on sales & marketing. Yahoo’s burn rate for those 9 months was $703k or $78k per month.
Was this round overpriced?
At approx. 20x-30x revenue run rate, the valuation seems challenging. But traffic was growing exponentially. It grew from 50k daily hits in Sep-94 to 1m daily hits in Jan-95 before reaching 6m daily hits in Mar-96.
This traction continued. Daily page views went from 6m in March 1996 to 20m at year end.
Yahoo IPO’d on 11-April-1996, only 4 months after its previous round of investment. Yahoo raised $35.0m equating to a market capitalisation of approx. $315m. Yahoo would go on to do $19.1m in revenue in 1996. This implied a forward revenue multiple of approx. 16.5x.
Was this IPO overpriced?
On its first day of trading, Yahoo’s stock jumped to $33 per share — almost 3x the IPO price. This resulted in a market cap of $866m and a forward revenue multiple of 45.4x.
Was 3x the IPO price overvalued?
Yahoo’s exponential traction kept continuing however. A year later in 1997, Yahoo reached 65m daily page views (4.6x growth); in 1998, it reached 167 million daily page views (2.6x growth).
Yahoo seemed to grow into its valuation, as it generated $67.4m in gross revenue in 1997. This growth implied a one year forward revenue multiple of 12.8x.
That being said, a fast growing revenue line should be unbundled to understand gross margin and marketing spend to achieve that revenue, before applying a multiple to revenue. On this measure, Yahoo generated $14m in gross margin after marketing. This indicates a forward gross margin after marketing multiple of 61x. This valuation feels challenging.
All of these rounds feel difficult to determine, even in hindsight, whether they were efficiently priced. Stepping back, the intrinsic value of any business is its future earning power and the discounted value of its forecast free cash flows. It is a range estimate rather than a precise figure.
Each year, management will seek to increase the business’s intrinsic value. This could be through improving its competitive position, increasing barriers to entry or increasing its pricing power for example.
A business’s market value is the market’s point estimate of intrinsic value at that time. The market, to paraphrase Ben Graham, is a voting machine in the short term. Ideally, market value will grow in line with the growth in intrinsic value. A bubble is a decoupling of market values from intrinsic values.
One could argue that Yahoo’s revenue growth is a proxy for growth in intrinsic value. This delta could then be mapped to associated change in market value. Yahoo’s revenue multiple would expand if growth in market value outpaced intrinsic value growth implying speculation.
For the period from IPO to 1997, Yahoo’s revenue multiple ranged between 6.9x to 16.0x. Its revenue multiple then expanded to 40.5x to 104.6x in 1999 (even as the business was at a much larger scale), before correcting in 2000. For the period following its trading multiple contracted to a revenue multiple of 3.3x to 7.0x. Using this revenue multiple proxy, Yahoo entered bubble territory during the period of 1998 to 2000.
Given today’s asking multiples, some might say that we are somewhere in the vicinity of 1997–98 on the eve of another bubble. Though bear in mind Yahoo is just one anecdote. And in fact from here all market scenarios are possible with an associated probability to each. Most would agree that it’s really hard to predict macro trends and the market’s response.
In my mind, another bubble could form if a new technology that redefined the landscape emerged. Something that allowed the market to say ‘this time it’s different’. Perhaps what may emerge from the blockchain will be that catalyst.
PS. Further reading
For further reading on Yahoo, I would recommend Marissa Mayer and the Fight to Save Yahoo.