Finding the Bottom: Are We There Yet?
What we know from recent downturns and the impact on venture capital returns.
Last week, the S&P 500 touched upon bear market territory by reaching an intraday low that was 20% off the January 2022 high. The Nasdaq composite is already in bear territory having shaved close to 30% off its all time high. This week the S&P 500 rallied to end a 7-week losing streak. Is this a sign that we’ve hit bottom? Who knows. But let’s see how previous downturns have turned out.
Recent Historical Perspective
How does the current pullback compare to previous downturns? The dotcom crash of 2000-2001 witnessed a 49.97% decline of the S&P from its high in March 2000 to the low in October 2002. The decline from peak to trough took a total of 31 months. The Great Financial Crisis of 2008 witnessed a 56.78% decline, but the decline was more accelerated having taken only 17 months from peak to trough. More recently, the Covid 2020 pullback was even more compressed with the S&P 500 declining 33.92% in a matter of one month.
While the current 18% downturn may feel apocalyptic, it is modest compared to other recent downturns. We are only 5 months into this correction, compared to the drawdowns that took 17 months in the GFC and 31 months after the dotcom crash. Given the rally this week, we are down only 13.31% from the January peak, so we are a far cry from the pullbacks we have experienced previously. As you can see from the following S&P 500 graph, we are nowhere near the magnitude we’ve seen from recessions of recent history.
But how much lower can it go from here? To get a sense, if we were to experience a decline similar to the 49% dotcom crash, the S&P 500 could fall another 41% from today. That would be even bigger than the crash in March of 2020. If we were to benchmark against the 2008 crisis, the S&P 500 could tumble another 50% from here. That would be like having a dotcom crash on top of what we have already experienced. The magnitude is hard to imagine, but it provides a better context from which to evaluate whether we’ve hit bottom.
Recovery
The duration for recovery from the dotcom crash and the GFC look remarkably consistent. The S&P 500 took 58 months from bottom to return to the highs prior to the dotcom crash and 50 months from bottom to recover from the GFC. The Covid 2020 shock, however, turned around in 5 short months. In tech, we observed that Covid-19 accelerated many trends, and apparently it accelerated economic cycles as well.
It’s an open question whether recovery of the current downturn looks more like the Covid-19 accelerated cycle or the pre-Covid recessions of recent history. What we do know is that the aggressive monetary and fiscal policies supporting the quick turnaround in March 2020, including the injection of unprecedented liquidity into the markets by the Federal Reserve, is no longer available given the Fed’s focus on controlling inflation. In addition to inflation, a number of additional issues pose further challenges to a quick 2020-style turnaround: rising interest rates, supply chain delays, rising unemployment, and uncertainties of war.
Impact on Venture Capital
How did venture capital as an asset class perform during the downturn and recovery period? While the extent of the correction and the duration of recovery were similar for the dotcom crash and the GFC, venture reacted very differently to the two recessions.
The following graph shows the performance of venture capital following the different downturns.1 To be fair, I’ve also included select indexes of public equities as well as fixed income to provide a point of comparison.2
It’s important to note the additional benchmarks are not offered as a way to evaluate the performance of deploying fixed capital into different scenarios. Asset allocators use public market equivalents, for example, to replicate the private market investment under public market conditions to generate how the investment would have performed in public markets. Here, I am just providing the actual historical performance by year without normalizing for timing of cash flows, distributions, etc.
Second, the IRR for venture capital represents the net IRR for median performance for vintages began in that year, whereas the public equities and fixed income benchmarks are the actual performance for that year. And of course, an allocator would expect a premium from venture capital due to the lack of liquidity when compared against its peers in more liquid markets.
That said, here are a few observations. As expected, you can see market volatility following the 2001 and 2008 periods, as reflected by the big swings of both the public equities and bond market during the recovery periods. In contrast, venture capital showed a steady rise through the period. Also, venture capital was consistently higher than one, if not both, of the other indexes throughout the period of volatility.
Venture capital benefits from the longer term investment horizons, so in order to smooth out the volatility of any particular year, the following takes the average over a period of time. Here you can see that the average IRR for venture in the 2000-2007 vintages following the dotcom crash hovered at a modest 6.68%. Because the dotcom crash was specifically precipitated and targeted at tech, that period still remains one of the lowest performing vintages for modern venture capital. However, the pubic market fared even worse with an average of 2.59% during the same period.
Venture returned to its heights following the GFC with an impressive average IRR of 19.28% for the vintages between 2008 and 2017. The top funds will be closer to 30% during this period. IRR is a less meaningful way to measure younger venture vintages, so it will take a few years to see more reliable performance for funds raised more recently.
Building for the Future
What accounts for the difference between the performance of venture following 2001 and 2008? For starters, 2001 was a crash precipitated by, and acutely targeting, technology companies. The GFC started in the financial sector, and although it ultimately spread to other sectors, technology was relatively insulated from the downturn. There were also two mega-trends that began at the time of the GFC: cloud and mobile. These trends continue to provide the tailwinds for technology companies today, and many speculate that we are on the cusp of new mega-trends in web3 and autonomy.
There are many theories why companies started in downturns have staying power. Many of the things that were hard during good times got a lot easier, such as the ability to attract talent. Also, these companies are forced to have financial and operational discipline, which itself create long term value for the company beyond surviving the immediate downturn.
But by far the biggest reason I believe venture keeps performing is that builders will continue to build. There is a rich tradition of technology companies being started following recessions, from HP in 1939 to Microsoft in 1975. Companies started closely on the heels of the GFC include Airbnb, Uber, Square, Whatsapp, Slack, Instagram, Pinterest, and Stripe. The opportunity cost for venture investors sitting on the sidelines and missing such as strong cohort was staggering.
For venture investors who continued to deploy capital following the last economic collapse in 2008, congratulations. They got in when valuations were in correction territory, and they gained valuable experience managing a portfolio during a downturn. The current downturn presents similar opportunities to reset valuations, develop skills to survive in lean environments, and participate in the next generation of category defining companies.
Builders will continue to build, and history has shown that founders and investors that support them have been vindicated in far worse conditions.
The VC IRRs are taken from this report: https://www.cambridgeassociates.com/wp-content/uploads/2020/02/WEB-2019-Q3-USVC-Benchmark-Book.pdf
To represent public equities, I’m using the S&P 500 composite index. To represent fixed income, I’m using the iShares 7-10 Year Treasury Bond ETF. Please let me know if there is a better benchmark to use to represent fixed income.