A perfect storm
The quantum technology sector faces a confluence of factors that will challenge its development over the next 2-3 years.
Tech market carnage
Quantum technology sees itself as a technology for the long term. Can’t it stand apart from short term market disturbances? Things are not so simple.
The global pandemic fuelled a stay-at-home tech boom. From Feb 2020, the Nasdaq composite increased by 65% to its Nov 2021 peak. These gains are rapidly unwinding. Carnage in the public tech markets inevitably puts pressure on private tech markets which are only just beginning to process this adjustment. After its own 2020-21 boom, the venture capital market is now in rapid retreat. Previous VC shocks have taken years to bottom out [1]. But this isn’t the only challenge.
Following the global financial crisis, a generation of entrepreneurs has come forward with little first-hand experience of non-zero bank base rates and positive real interest rates. The after effects of public pandemic spending, deglobalisation and the continuing war in Ukraine mean that inflation now once again stalks the globe. An extended period of much higher real interest rates will likely be required to get it back under control. [2]
Why does this matter? The real cost of capital is going up and it’s likely to stay up because that is actually normal. There may even be a real crunch. Investors are again being offered more choices for how their capital can earn a return. Deep tech famously looks to ‘patient capital’. However, this is a commitment to a time horizon, not to charity.
Lengthening revenue timelines
It’s long been clear that the best known quantum applications require large, high-fidelity machines, and those are still some way off. In 2017, John Preskill asked the question, what might be achievable with noisy intermediate scale quantum (NISQ) devices? What would have to wait for large scale fault-tolerant quantum computation (FTQC)?
The NISQ era – on Preskill’s original definition a NISQ device should not be classically simulable. In 2019, Google Sycamore demonstrated the first ‘beyond classical’ calculation (on an artificial problem) with increasingly convincing demonstrations following in 2020 and 2021 from the USTC Jiuzhang and Zuchongzhi devices to follow. Most recently Xanadu’s Borealis promises to make the execution of such jobs routine and available to cloud based users. Finally, the wider community can experiment with such calculations directly (though GBS is still an artificial problem, with no known practical application) [3].
Even as we finally enter the true NISQ era, expectations are narrowing. It’s always been clear that such devices could not run full scale quantum algorithms (e.g. Shor’s algorithm for factoring, Grover’s search algorithm or HHL for linear algebra). However back in 2017 it was reasonable to argue that a hybrid quantum-classical approach using new variational quantum algorithms might offer a route to quantum advantage in a wide variety of practical applications. We now understand much more clearly why at least the basic implementations of these techniques, are not likely to lead to a broadly realisable economic advantage.
Variational techniques continue to develop. Hybrid architectures continue to become more efficient. Other techniques such as analogue quantum simulation offer their own promise. Some early applications (such as certified random numbers) are real but relatively niche. Direct experimentation with genuine NISQ devices may unlock new opportunities. There are certainly grounds for optimism that such devices will continue to be useful for advancing quantum science itself. [4] Quantum annealing continues to hold promise. In a notable milestone, optimising crane lifts at the Port of Los Angeles is an example of a deployment into routine commercial use [5]. However, such implementations don’t necessarily show that QA can sustain an advantage over the best classical algorithm, or fend off the challenge from quantum-inspired implementations [6]. A similar story is set to play out in financial portfolio optimisation.
We still don’t know if a more transformative commercial quantum opportunity will emerge in the NISQ era. However, it’s difficult to avoid the conclusion that the prospects for this are currently receding versus previous market expectations.
Other sectors of quantum technology offer alternative routes to realising benefits, but face their own challenges.
Quantum communications offers its own early opportunities. QRNG and QKD devices are already at market. QRNG is a unique capability, but faces a crowded and distracted marketplace. QKD has unique strengths, but has to sell into the teeth of a maths-based security incumbency. NIST’s announcement of its selections for PQC standardisation will draw further airtime towards such solutions. [7] Quantum sensors offer diverse opportunities and much potential, but their market is fragmented. In each niche these new comers typically face mature and well-engineered incumbent solutions.
Unwinding hype
A sharply deteriorating capital market prognosis and lengthening roadmaps to revenue are not the only issues. These fall on a sector that was anyway right at the top of its hype cycle.
Interest in the advanced applications of quantum theory originally grew in the academic community. Later it was championed by academic groups sponsored by tech majors. Accelerating from 2014 these have been followed by a growing number of specialist quantum startups. Over the last couple of years this has spawned quantum projects within a rapidly growing community of corporate would-be end-user. In this year’s quantum conference season, it’s been noticeable that quantum enthusiasm has been carried by this latter cohort; just at a time when early academic proponents are calling for a break on timescale expectations.
SPAC – As they grow and their capital needs increase, most businesses are ultimately attracted to the liquidity and deep pockets offered by public markets. But how to implement a listing? In the 2020-21 boom SPACs became the fashionable route of choice, circumventing much of bureaucracy (safe guards) of the traditional IPO route [8]. Unfortunately, such a process can act to reward rather than temper hype, at least in the short term. Disgruntled investors and class action law suits have been the predictable result across the sectors where SPAC deals have been popular. Quantum-sector SPACs are no exception [9].
Just as other pressures build on the wider sector, two would-be quantum champions IonQ and Arqit face the prospect of having to keep one-eye on SPAC-related class action litigation.
The course remains true
It’s important to realise that the long-term quantum journey remains strongly on track.
Underlying progress is strong
For some, the growing tempest will lead to an adaptation of the classic trope ‘useful quantum computing is 10 years away and always will be’. However, this is not where the evidence points.
For a recent review of progress see Quantum Outlook 2022.
We shouldn’t be equally impressed by all of the different technology variations and roadmaps on display. But overall progress towards the field’s longer term FTQC goals has been robust. Fidelities have improved across a range of modalities. Logical qubit precursors have been demonstrated by multiple groups. Fabrication innovation is opening up a new fronts on scalability. Innovation on quantum error correcting codes is a tantalising prospect. [10]
Killer apps await
It’s sometimes said that quantum computing is still looking to nail its killer app. This is misleading. Quantum computing already has not one but two epoch-defining killer apps: breaking current cryptography (an inevitable geopolitical driver); Quantum chemistry & materials science (foundational for pharma, wider biotech and advanced manufacture). It no longer seems a question of if machines large enough to deliver these applications will be developed, but when, where and by whom.
An additional killer app still awaits to be truly nailed. This is based on the intuition that QC should also be an important accelerator for machine learning and AI. Some of the things we need here are less in the bag. We’ll need QRAM to enable a full range of linear algebra speedups. We’ll need faster QPU architectures before we can routinely exploit general purpose optimisation speedups. Progress is intriguing: quantum machine learning on uniquely quantum data; quantum natural language processing. [4]
Wider quantum technology is set to bring its own big-ticket benefits to the party. Quantum networks promise to be a unique exponential multiplier for the power of quantum computing. Quantum sensors offer foundational sensitivity and, in some cases, ‘inherent’ calibration. More speculatively, their own killer app may be their use in combination with quantum machine learning.
The deepest of deep tech
Long-term economic growth and so commercial opportunity is ultimately driven by technological innovation.
In Economic Transformations Lipsey, Carlaw and Bekar argue that long-term economic growth in the West has been driven by a series of technological revolutions. In particular they identify the development and adoption of a series of ‘general purpose technologies’ as key growth drivers. In the modern era these include printing, the steam engine, electricity, the computer and others. This focus naturally leads to an emphasis on the role of science in supporting sustained growth and sustained national competitive advantage. [11]
Quantum isn’t just a new technology. The advent of quantum mechanics has been one of our most far-reaching scientific revolutions. Literally, not just metaphorical. In modern times only evolution, relativity, molecular biology and complexity theory come close in their impact on our basic scientific thinking.
To see that quantum technology is another key general purpose technology is to realise how wide spread its spill-over impacts are set to be. Push hard enough in almost any area of the natural sciences and you ultimately find yourself rubbing up against quantum effects that need to be modelled. More generally, quantum information science seeks to push even further into the foundations of computability.
In general, the intersection of frontier science and next generation technology is known as deep tech. Many see developments here as being key to addressing many of the world’s most pressing problems: clean growth, ageing populations, social equity and improved governance [12]–[14].
Quantum technology can expect to act as an accelerator for other deep tech segments, including AI, biotech, robotics, advanced materials, blockchain and space. Ultimately, the most influential future businesses may be those that bring these technologies to market in combination.
Meeting the challenge
Quantum startups, existing investors and governments all face their own challenges in riding out the storm.
Quantum startups
The wider quantum technology sector already has seen over 580 startups. Most of these have not yet battened down for the heavy weather ahead.
Many quantum hardware players are committed to expensive R&D programmes, particularly so where fundamental scaling challenges remain to be solved. Quantum talent has long been flagged as a key bottleneck. Quantum software and algorithm startups in particular have already experienced high rates of salary inflation.
The storm has caught many in the quantum flotilla unawares. Some are fortunate, having raised substantial funds recently during the boom. Others will now need to head out directly into the wind.
Creative destruction – a common defence of market corrections is that they ensure that capital is going to the most deserving businesses. But how sure are we that the quantum startups that happen to be sitting on boom-time funding are the most deserving? This is not at all clear. Early academic priorities in the quantum sector were typically to develop basic platforms that could prove key concepts, such as the basic demonstration of ‘beyond classical’ calculations. Commercial investment has tended to seize on these ‘devices’ as a tangible focus for investment. However, in the absence of significant NISQ revenues, a practical roadmap for scaling-up could be much more important. Most qubit modalities haven’t really set-out on this journey yet.
Very early-stage startups should still be able to access modest amount of seed capital. However all startups, and particularly high spending quantum startups, now need to carefully watch their cash reserves and defensively plan towards the next point at which they will raise capital. Some experienced hands are warning their portfolio companies to be ready to wait to 2025 if they can [15]. How easy this will be to achieve varies widely across market segments.
Hardware companies with expensive development roadmaps face particular challenges. Typically, they will have planned their funding around technical milestones (with an implied significant valuation increase at each stage). Fact Based Insight believes the market hasn’t paid enough attention to the actual R&D costs specific to individual approaches (the details vary wildly even within qubit modalities). New startups will face tough questioning because the table stakes are already high. However, Fact Based Insight believes these remain merited where they can demonstrate unique advantages in meeting fidelity or scaling challenges. The right development partners will be essential. Software companies need to be realistic and robust about the sources of revenue they are seeking to access. A culture of nimbleness will likely be essential. All need to carefully watch their salary bill and have appropriate tools in place to allow its management. Fact Based Insight believes software startups can still target NISQ era value, but they need to be specific in their plan of how this will be achieved. Quantum-inspired solutions are a legitimate part of the mix. Startups further up the value chain, will often be able to access ‘picks and shovels’ opportunities to bolster near-term revenues. Startups focussing on opportunities in cybersecurity, timing, sensing or imaging can target their own near-term revenues. Fact Based Insight continues to see these as attractive areas for those dissuaded by the long horizons, high capital requirements and competitive intensity of the pure quantum computing segment.
Existing investors
We might expect VC to naturally turn market turmoil into an opportunity, however there are real practical challenges. Recall some of the basic mechanics of how VC actually works: limited partners (the ultimate end-investors) commit money to general partners (managers of venture funds) to invest in a portfolio of businesses. The GPs have executive control, but the LP’s money is only called down at the time investments are actually made. At a time when LP’s wider portfolios are under pressure, they won’t welcome GPs rushing to call down funds unless a reset in valuations throws up compelling opportunities.
Mark downs in public tech markets are already spilling over into private market valuations. This is a double-edged sword for GP’s. Their existing portfolio investments may now be sitting under water, further tempting the wrath of their LPs.
A key reality for many quantum businesses is that their road to significant revenues may be a long one. To bring quantum visions to realisation, particularly where hardware development is required may take hundreds of millions of dollars; where fundamental scaling challenges remain to be solved perhaps billions of dollars. Even the best funded in the current generation of startups probably don’t have the firepower to go all the way.
Early investors need to manage three key questions. Will a deep pocketed source of funds to support continued investment be available? Do they have funds to follow-up with their own part in these investments, or can they accept the dilution of their equity holding that will ensue? Crucially, can the valuation of the business continue to be managed smoothly upwards?
In The Secrets of Sand Hill Road Scott Kupor argues that venture capital funds should eschew chasing a normal portfolio of conventional business returns. Instead they should focus on generating a few ‘home run’ investments that grow spectacularly in value, offsetting losses elsewhere. [16]
In the aftermath of the storm, venture capitalists may worry about a dearth of new capital entering the market, they may want to reserve even more capital to follow-on their existing investments.
Might corporate capital pick up the slack? The R&D divisions of major defence, aerospace, energy and telecoms companies have invested directly in their own deep tech since before the term was coined. More recently many have set-up teams pursuing strategic investments along VC lines. There is very much an industrial logic for many corporates to invest long-term in quantum. We are likely to see direct support for startups via collaborations and proof-of-concept projects continue to grow. However new corporate VCs will need allies in a market under pressure.
Governments
Why should governments care? With its long-term promise still fully intact, the quantum sector remains one of the key competitive arenas for 21st century economies. Its also a unique opportunity for a reset on digital tech: Silicon Valley is a player in the race, but it’s not the only possible home base for this new technology. Governments around the world have already been active in staking their claim to these emerging economic high grounds. The desire to do this is unlikely to change. The means may come under pressure.
Government interest in quantum is part of a much wider revival of interest in active industrial policy. A key objective is to promote their country or region as the place where innovative ideas can be nurtured and grown into major businesses, and importantly, into clusters of major businesses.
In The Entrepreneurial State, Mariana Mazzucato pushes back on the notion that industrial policy is just a doomed bureaucratic attempt to pick winners. Instead, she analyses the role of the US state and programmes such as ARPA (later DARPA) in launching the current era of US digital tech leadership. In particular she highlights the importance of sovereign demand and R&D spending in supporting the early formation of Silicon Valley. [17]
A key part of this challenge is often seen as how to help innovative startups bridge the ‘valley of death’, the period when they need to burn cash to complete necessary R&D and to grow but aren’t yet generating revenues to underpin these costs.
In Boulevard of Broken Dreams, Josh Lerner surveys lessons learnt from more recent government attempts to stimulate entrepreneurship and venture capital. He explores the pitfalls from subsidising incumbent giants, to capture by vested interests. He also opinions on what it takes to get things right, from the legal framework, to training for entrepreneurs and access to cutting-edge technical infrastructure. Successful venture case studies include Singapore and Israel. [18] In China capital funding has also been a key emphasis. Between 2015-21 the government has raised almost $1 trillion in around 2000 ‘guidance funds’. These are designed to allow the flow of government support to startup businesses to be guided by private capital participation. [19]
In short, much government thinking around the world has focussed on how to get academic ideas out of the lab and into a self-sustaining and virtuous capital markets funnel. This is a key pillar on which they plan to build their future tech ecosystems. How to channel seed funding to the right aspiring startups? How to serve the larger capital needs of major growth businesses? How to offer attractive public markets as a long-term home for large businesses?
The current market storm threatens to swamp existing programmes. Explicit or implicit government support is increasingly important in securing follow-on commercial capital. Many governments will feel pressure for increased involvement to capture growth for their ecosystem.
Quantum sector startups have often been based on local academic expertise. Local government quantum programmes mean that in many cases these have been able to take advantage of grant funding and ‘soft’ contract revenues. Such incentives always face a danger of ‘capture’ by vested interests, which is why there is typically a desire to blend such support with participation from commercial investors and demand-side pull from potential customers. In the coming period, public policy makers may need to do more than they would like to keep activity moving forward.
Medium term growth funding is also a key concern. Quantum sector hardware businesses in particular are likely to need very significant funding. This is beyond what conventional tech VC funds can support, particularly outside of North America. In Europe, large institutional investors have typically avoided such high-risk segments. How, and how safely can this be changed? Pension funds investing in growth businesses doesn’t lower the risk profile of these segments.
In the long-term, mature businesses will typically seek the depth of liquidity and capital cost efficiency that public markets provide. The difficulties involved in attracting successful businesses away from established US markets is only too well illustrated by the current battle over the location of ARM’s forthcoming listing. The controversy over such policy ‘even amongst allies’ betrays the tensions that exist over both the philosophy and implementation of industrial policy [20].
Finding opportunity
Quantum has much to offer. A great technology with unmatched foundational potential. Smart founders eager to drive it forward. Governments keen to offer support. Better still (for new investors) a coming reset in valuations. For those with the right funding and right approach this represents a great opportunity.
As always, the ultimate questions in investment are ones of selection, valuation and timing.
Selection
Quantum technology is daunting for the uninitiated. There is naturally a temptation to compartmentalise these complications. But how safe is it to bring over business assumptions from the conventional digital tech stack?
Traditional digital markets are based on underlying technology that is relatively well proven. There is little true technical risk, instead market risk and business risk dominate. VCs therefore typically put primary emphasis onto evaluating the strength of the team taking the business to market. In Quantum the team still matters, but technical risk is too important to ignore. Traditional digital tech investors often prefer software businesses because they are perceived as offering a quicker route to market. Obviously, that’s not so simple in a sector where the hardware isn’t ready. Investors also see software as a hardware-agnostic hedge. While there is truth in this, it’s not a sure thing. An expensive staff salary bill needs to be financed. The circuit model of quantum computing is not the only way things may play out.
The quantum value chain is different. The unique aspects of quantum greatly affect the importance and dynamics of particular value chain steps. This isn’t just about due diligence that drills narrowly down into a tech. Selecting the right investment opportunities requires an understanding of potential bottlenecks upstream and downstream of the specific business case. Due diligence needs to be conducted ‘heads-up’ in a wide value-chain context.
Valuation
We aren’t close to being able to value quantum businesses on a conventional cash-flow basis, nor even on benchmark multiples of revenue or some other hard metric. We’ll usually be forced down into the ‘art’ of venture style valuation.
It’s common place to say that ‘we don’t know yet which hardware technology will win-out in the long-term’. Indeed, this may be true in a general sense especially when we consider diverse niche applications across the wider sector. However, this hides the clear differences that emerge when we examine specific technical programmes in more detail. Within each modality, individual companies differ radically in their approach.
A technically informed view is necessary. Fact Based Insight believes that current market pricing has been very inefficient in reflecting the different individual roadmap risk profiles that individual players face in bringing their technology to market.
Timing
As with any period of turmoil, the best timing is to spot opportunities as they emerge in advance of the wider market. Fact Based Insight see three themes that it believes are worth paying attention to.
Second wave founders – The first wave of quantum startups has mostly been formed by academics making their first move outside academia. The history of Silicon Valley would predict that an important second generation of startups form when engineers disillusioned with progress at early startups jump ship and set-up their own outfits. Will this work in the rarefied quantum talent market? Where will it be practical to build a new team? Will government programmes find it easy to nurture and support such projects? Seed capital may find a profitable home but will have to be discerning. Distressed gems – In conventional markets we often ascribe the failure of startups to underlying weaknesses in their business models and the inability of their founding teams to turn ideas into paying customers. Some quantum startups will deserve to fail for exactly these reasons. Others will contain hidden technological gems. Some have just been caught-out by the storm. Smart investors should be equipped to spot these and be ready to invest to put them on the right path forward. The right stuff – At the larger end of the market, quantum leaders will start to accelerate their ambitions to roll-up the sector. Such an approach requires an open technical vision, one able to plug in and add value to a wide range of partners, strategic allies and ultimately acquisitions. It also requires the right company culture, one able to incorporate the energy and new ideas coming from acquisitions. One able to channel star talent while keeping immediate salary and resource expectations under control. Unless you are already a tech major, public markets are a difficult place to conduct such a journey. Such businesses need the right growth funding partners.
The quantum journey continues. Key technological developments will likely continue over the next 10 years and beyond. However, commercially the next 2-3 years will be a key challenge. The startups, investors and governments who best seize the unique opportunities presented will be well set to make 21st century history.
References
[1] ‘Venture capital’s reckoning’, The Economist. Accessed: Jul. 02, 2022. [Online]. Available: https://www.economist.com/leaders/2022/06/30/venture-capitals-reckoning
[2] S. Frank, ‘Real Interest Rate – 151 Year Chart | Longtermtrends’. https://www.longtermtrends.net/real-interest-rate/ (accessed Jul. 03, 2022).
[3] L. S. Madsen et al., ‘Quantum computational advantage with a programmable photonic processor’, Nature, vol. 606, no. 7912, Art. no. 7912, Jun. 2022, doi: 10.1038/s41586-022-04725-x.
[4] ‘Quantum Algorithms Outlook 2022’, Fact Based Insight, Dec. 17, 2021. https://www.factbasedinsight.com/quantum-algorithms-outlook-2022/ (accessed Jul. 08, 2022).
[5] D-Wave Systems, Logistics Optimization with Quantum: Port of Los Angeles, (May 18, 2022). Accessed: Jun. 29, 2022. [Online Video]. Available: https://www.youtube.com/watch?v=jA7iopqDm94
[6] H. Oshiyama and M. Ohzeki, ‘Benchmark of quantum-inspired heuristic solvers for quadratic unconstrained binary optimization’, Sci Rep, vol. 12, no. 1, Art. no. 1, Feb. 2022, doi: 10.1038/s41598-022-06070-5.
[7] ‘Quantum Internet Outlook 2022’, Fact Based Insight, Dec. 15, 2021. https://www.factbasedinsight.com/quantum-internet-outlook-2022/ (accessed Jul. 08, 2022).
[8] ‘How Spacs went splat’, Jul. 06, 2022. https://www.ft.com/content/1c13119f-d08b-4024-9f90-da997add0582 (accessed Jul. 08, 2022).
[9] ‘SPAC Task Force’. https://www.rgrdlaw.com/services-litigation-spac-blank-check-task-force.html (accessed Jun. 29, 2022).
[10] ‘Quantum Hardware Outlook 2022’, Fact Based Insight, Dec. 13, 2021. https://www.factbasedinsight.com/quantum-hardware-outlook-2022/ (accessed Feb. 02, 2022).
[11] R. G. Lipsey, K. I. Carlaw, and C. T. Bekar, Economic Transformations: General Purpose Technologies and Long Term Economic Growth, Illustrated edition. Oxford ; New York: Oxford University Press, USA, 2006.
[12] ‘Moonshot R&D|About’. https://www.jst.go.jp/moonshot/en/about.html (accessed Jul. 08, 2022).
[13] ‘The Grand Challenge missions’, GOV.UK. https://www.gov.uk/government/publications/industrial-strategy-the-grand-challenges/missions (accessed Jul. 08, 2022).
[14] ‘EU Missions in Horizon Europe’, European Commission – European Commission. https://ec.europa.eu/info/research-and-innovation/funding/funding-opportunities/funding-programmes-and-open-calls/horizon-europe/eu-missions-horizon-europe_en (accessed Jul. 08, 2022).
[15] ‘The great Silicon Valley shake-out’, The Economist. Accessed: Jul. 04, 2022. [Online]. Available: https://www.economist.com/business/2022/06/28/the-great-silicon-valley-shake-out
[16] S. Kupor and E. Ries, Secrets of Sand Hill Road: Venture Capital—and How to Get It. Virgin Digital, 2019.
[17] ‘The Entrepreneurial State: Debunking Public vs. Private Sector Myths eBook : Mazzucato, Mariana: Amazon.co.uk: Books’. https://www.amazon.co.uk/Entrepreneurial-State-Debunking-Public-Private-ebook/dp/B0759WNBVC/ref=sr_1_1?crid=260CYY2FUFSIE&keywords=the+entrepreneurial+state&qid=1656667103&s=digital-text&sprefix=the+entreprenurial+state%2Cdigital-text%2C56&sr=1-1 (accessed Jul. 01, 2022).
[18] J. Lerner, Boulevard of Broken Dreams: Why Public Efforts to Boost Entrepreneurship and Venture Capital Have Failed–and What to Do about It. Princeton University Press, 2009.
[19] ‘The rise of China’s VC-industrial complex’, The Economist. Accessed: Jul. 01, 2022. [Online]. Available: https://www.economist.com/business/2022/06/27/the-rise-of-chinas-vc-industrial-complex
[20] ‘UK’s Arm-twisting confirms critics’ worst deal-vetting fears’, Financial Times, Jun. 24, 2022.
Comments