Something ironic and seemingly paradoxical is happening in the world of capital investment.
Just as money got expensive, with interest rates at their highest levels in 15 years, funds and companies are pouring unprecedented sums into massive infrastructure projects. Trillion-dollar investments in AI infrastructure. New chip factories. EV gigafactories. Energy infrastructure buildouts. VCs backing capital-intensive ‘American Dynamism’ moonshots like supersonic planes and defence tech.
This wasn’t supposed to happen. The conventional wisdom said these ambitious, capital-heavy projects should have flourished during the 2009-2021 era of zero interest rates – when money was essentially free. Instead, that era gave us primarily asset-light software plays: SAAS companies, marketplaces, and apps like Uber and Airbnb that merely ‘unlocked value’ in existing infrastructure rather than building anything new.
Why are we seeing an explosion of ambitious infrastructure investment now, exactly when borrowing costs are higher? I have a theory about this apparent paradox – one that challenges how we think about the relationship between interest rates and progress.
The conventional logic: Low rates extend time horizons, making bets with longer term payoffs more valuable, allowing innovative projects to get funded.
I am currently reading the excellent Boom: Bubbles and the End of Stagnation, and I was taken by how the authors seem to tie themselves up in knots trying to explain how lower rates in a fiat financial system were continuing to cause stagnation.
“The impact of low rates can sometimes be paradoxical. In one sense, they’re a sign that the supply of money exceeds the demand for applying it to interesting problems… It’s tempting to say that low rates actually promote investment in the sorts of productive projects the world needs. The trouble is that low rates indiscriminately encourage long-term projects, while the biggest episodes of wealth creation come from building on a single theme. A low-rate environment does make a fusion power company or an experimental anti–aging drug more fundable, but it also means that those companies are competing with every other project with a potential long-term payoff.”
I agree with the Boom authors Byrne Hobart and Tobias Huber conceptually, but I think their framing is a confusing way to think about the issue. They argue that lower rates cause everything to get funded, and this is not ideal because it doesn’t allow for the focus on a specific sector. I get the point they are trying to make, especially in relation to their thesis that sector specific bubbles are not only not bad but actually help drive progress.
But I think there are different ways of looking at this situation that may help us understand the “paradox” or irony.
The inverted logic: lower rates lower the hurdle rate for outcomes, resulting in worse results and returns
“Invert, always invert” – Charlie Munger
While it’s true that lower interest rates extend time horizons, they also tend to lower cost of capital which implicitly tends to lower the acceptable “hurdle rate” for projects to get funding. Inverting the formula would suggest that higher interest rates therefore increase the hurdle rate for projects to receive funding.
Alex Danco wrote in Positional Scarcity that “In conditions of abundance, relative position matters a great deal”. Packy from Not Boring recently used that quote while discussing AI’s potential to liberate humans from economic scarcity, but I am now thinking about it in relation to the ZIRP era. In conditions of an abundance of capital, the relative positioning and incentives of capital allocators (VCs) and capital deployers (founders and startups) changed.
Either intuitively or in a calculated manner, capital allocators (VCs) and capital deployers (founders and startups) adapted to the new zero interest rate market environment, and the relative positioning similarly shifted. Return on Invested Capital doesn’t matter if capital is abundant. “Attention” became the new scarce resource, and “winning the market at all costs” was the goal, the “at all costs” meaning the deployment of the “capital cannons” from Softbank, Tiger Global, or any of the other funds that raised and deployed tens of billions of dollars in that era.
Return on Invested Capital isn’t just an outcome. It’s a function of the utility created by the capital deployer, or how productively they deployed the capital. ROIC is a useful gauge on the value a capital deployer is creating.
And when money is effectively free for the founders deploying capital, the competitive vector shifts to who has access to the biggest cannon in order to do more, with more. This might build big businesses, but isn’t truly innovative. WeWork built a big business by deploying a lot of capital, not because it was truly building an innovative technology to increase society’s productive output or utility. WeWork deployed billions and billions in capital, and remains a relatively large business even today, post-bankruptcy. But the ROIC on those invested billions was negative. WeWork won the “attention” game in the ZIRP era, but didn’t create any lasting value.
So having an extremely low cost of capital creates perverse incentives that lead capital allocators and deployers alike to play games that aren’t directly related to creating true value, and therefore low rates can produce worse outcomes.
But is there a case to be made for why higher interest rates could actively push us towards more progress?
Why would higher rates drive progress?
One of the things I really like about Boom is the distinction they make between innovation, a word which tends to lean more towards software and is overused in modern society, and progress.
Innovation is optimized workflows, automated backup to the cloud, and simplified systems integrations. But progress is the future we and past generations imagined. Flying cars, abundant energy, personalized, affordable health care, and all the rest.
I have a personal anecdote that might help understand why higher rates might lend itself more towards progress than innovation.
I work at Somos Internet, a next-gen vertically integrated ISP bringing data-center network architecture to entire cities, building active ethernet fiber networks instead of passive fiber telco networks that were designed for a world where compute was still the scarce resource. We have developed a model whereby we are connecting any new “homespassed” ie. a house or apartment that becomes active for coverage in our network for less than $10 of capex. This is an order of magnitude better than the local competition who are quoted as spending at least $100 x passing to build a lower quality network. Our cost to build a network is so cheap that we can achieve payback on the activation of a customer in 6-12 months vs. 3+ years for other telcos.
Somos is just getting started and needs capital to scale. We had raised a seed round following YC, but ideally we want to fund capital expenditures with debt, not equity. So in 2024 we signed a $30mm Forward Flow financing agreement with a US based Private Credit fund, unlocking capital as the company originates new receivables to be reinvested in network growth. The rates on this facility are not cheap (what you might expect for a negative cash flow startup based in Colombia).
But crucially, the rates are acceptable for the company given the incredibly attractive underlying unit economics of the business that the facility is financing. We can make the economics work because we will be deploying that capital in an incredibly productive manner.
Cost of capital is a huge weapon that favors the scale incumbents vs. smaller upstarts.
This spread over the risk free rates applies whether in a low interest rate environment or a high rate environment. But interestingly and importantly, the relative % of the spread in credit risk between incumbents and upstarts is lower in a high rate environment than a low rate environment.
An example: With SOFR at 1%, an incumbent at SOFR + 3% and an upstart at SOFR + 7%, the relative difference of the cost of capital is double (the incumbent will pay 4% vs. the upstart paying 8%, 100% more), but in a world where SOFR is at 5%, the relative difference in the cost of capital is materially smaller (the incumbent will pay 8% vs. the upstart pays 12%, only 50% more).
Higher interest rates therefore not only remove perverse incentives from zero interest rate environments, but actively serve to reduce the relative cost of capital disadvantage faced by genuine new technologies and approaches that can increase utility and productivity.
A different example outside the world of finance can further illustrate the argument:
When looked at abstractly, the cost of labor vs. capital is a similar situation. In industries as diverse as clothes manufacturing and call centers, a low cost of labor disincentivizes investment in technological solutions. Sure, they might use software, but it is still millions and millions of low cost humans employed to attend to the industry’s needs. Cheap labor is the “incumbent” and while it remains marginal to the cost structure, there is little incentive to apply technology. However, as the cost of labor rises (akin to a higher rate environment), the relative advantage of the “incumbent” technology, in this case people’s manual labor, is diminished.
With the adoption and ongoing innovation in AI, we are seeing progress as a new technological solution is providing utility that far exceeds the status quo, even despite materially higher capital costs. AI is so much better that the relative cost of capital doesn’t matter.
Had labor been “effectively free”, even AI might not have displaced it. But the Cost of Labor represents a sufficiently high hurdle rate that progress occurs when a new solution has such utility that it supersedes the relative cost of capital disadvantage it faces.
In a high interest rate environment, the cost of capital goes up for everyone, but the upstart’s relative disadvantage is reduced.
If an upstart like Somos has an underlying unit economic profile that is materially differentiated to the competition, Somos can actually more easily overcome the cost of capital advantage held by the larger incumbents in a higher rate environment than in a lower rate environment. Effectively the higher hurdle rate in the higher rate environment makes it harder for the incumbent to mitigate the underlying unit economic advantage held by Somos.
Progress happens when the increased utility or productivity of a new technology or output far exceeds not only an incumbent company’s hurdle rate, but also manages to exceed the hurdle rate of a new company with a cost of capital disadvantage.
So in a high interest rate environment where everybody’s cost of capital is higher, the only thing that can possibly get funded and can possibly thrive is a step change innovation with a hurdle rate far in excess of the status quo, even with high costs of capital. And in that environment, that project or company actually stands a better chance of succeeding vs. incumbents than it would in a low interest rate environment.
There seems to be a consensus that the investments being made today are different to the ZIRP era WeWork-style of capital deployment. That these investments are dollars being put to work productively to generate new utility that exceeds the status quo. I think I fall into that camp, too.
If that is the case, what we are now seeing is capital-p Progress, not only unencumbered but higher interest rates, but potentially as a direct result of them.
As I have noodled on this idea over the past 6 months or so, I have also spent a long time thinking about the Fed’s dual mandate of Maximum Employment and Price Stability, and the implicit third mandate since the GFC of avoiding financially driven recessions. The reduction in volatility of both employment and prices since the 70s has served to pacify the public, but it has likely served to entrench the status quo. What price have we paid in the form of foregone progress?
Thanks to Claude.ai for help editing.
Ross Rambles 