This note captures some reflections after a summer internship in quantitative finance. My goal here isn't to argue that the financial industry lacks value, but rather to untangle a more personal question: in what ways does finance genuinely create value, and is that value the kind of thing I want to build a career around?
Before the internship, my view of finance was fairly textbook: markets need liquidity, they need price discovery, and they need mechanisms to transfer risk from those who don't want to hold it to those who are willing to. Market makers allow people to buy and sell on demand; researchers and investors aggregate dispersed information into prices; derivatives and financing tools help businesses, pension funds, and individuals manage uncertainty.
After seeing it up close, I still don't dispute any of those functions. What I realized, though, is that they don't fully answer the question I care most about: to what extent does the day-to-day work meaningfully improve people's lives or the real economy, and to what extent is it about competing for an edge within a system that already works remarkably well?
I now tend to think of finance in two distinct layers. The first is its foundational infrastructure: capital formation, risk management, clearing and settlement, baseline liquidity, and long-term capital allocation. The second is the layer of continuous optimization built on top of it—faster quoting, tighter spreads, more intricate structures, and ever-finer information processing. The first layer is undeniably essential. Where I find myself pausing is on the second: whether its marginal impact always matches the extraordinary amount of talent, capital, and attention it draws.
I used to hold a somewhat idealized view of "price discovery," as if market prices steadily converge toward some objective, underlying truth. Now, I'm more inclined to see prices as a reflection of the collective consensus of market participants at a given moment—their shared estimate of future cash flows, risks, timelines, and uncertainties, and the price at which they are willing to trade on those beliefs.
That consensus certainly absorbs real information and is often sharper than any single person's judgment. Still, it isn't reality itself, nor is it the final verdict on the worth of a technology or a company. Who holds the capital, whose voice carries weight, and who faces leverage or time constraints all shape the price on the screen.
The early 2025 swing in quantum computing stocks is a good illustration of this. Driven by technical milestones, broader AI enthusiasm, and long-term optimism, those stocks had been climbing steadily. Yet in many industry discussions at the time, practitioners remained quite measured about those valuations; how long it will take to build truly large-scale, fault-tolerant, and commercially viable quantum computers is still a wide-open question.
In early January, Jensen Huang remarked that "very useful" quantum computers were likely still 15 to 30 years away—noting that 15 years might be on the early side, 20 years a timeline many would believe, and 30 years perhaps on the conservative end. Related stocks sold off sharply. What struck me as interesting is that, to many researchers actually working in the field, a 15-to-20-year horizon for truly transformative, large-scale quantum computing isn't necessarily pessimistic—some might even view it as a fairly encouraging long-term outlook. Nor is Huang a quantum computing researcher himself. Yet the market didn't read the comment as "the long-term future looks promising"; it read it as "commercialization is much further off than priced in."
In that sense, the price movement wasn't necessarily discovering the intrinsic value of quantum computing. It was repricing a narrative the market had already bought into—recalibrating timelines, probabilities, and discount rates. Market prices aren't mere noise; they encode information, expectations, and risk appetite. Primarily, though, they tell us how the market interprets the world, not necessarily what things are truly worth.
I still believe liquidity has real, tangible value. Corporations, funds, pensions, and everyday investors all need to transact. When assets are hard to trade, investors demand higher returns, and companies face higher costs of capital. Market-making, clearing, and execution infrastructure make all of that cheaper and smoother.
At the same time, liquidity isn't free. Someone has to carry inventory risk, absorb adverse selection, and weather extreme volatility; someone has to step in and buy when others are rushing to sell, or sell when others are eager to buy. In calm markets, this feels effortless and invisible. In stressed markets, the real question surfaces: who still has the balance sheet and willingness to absorb risk?
When that happens, market makers, hedge funds, investment banks, asset managers, commercial banks, insurers, and clearinghouses all respond in different ways. Some widen quotes, deleverage, or unwind positions; others with more stable capital step in to buy forced sales; some arrange financing or hedging for businesses; and some, through their own leverage and counterparty links, end up transmitting stress further.
It's hard to neatly label these institutions as purely "stabilizing" or "destabilizing." Financial systems can both absorb shocks and amplify them. The difference usually comes down to capital adequacy, funding stability, risk concentration, and whether everyone is forced into the same trade at the same moment.
There's no question that a liquid, transparent, low-cost market is vastly better than one with wide spreads, scarce financing, and unmanageable risk. The subtler question is: once a market is already highly efficient, how much is the next sliver of optimization really worth?
Take heavily traded U.S. equities. Under normal conditions, spreads are already razor-thin. For a $100 stock with a $0.01 bid-ask spread, the full spread is just 1 basis point (0.01%). Measured against the midpoint, the immediate cost to buy is roughly half a basis point, and a full round-trip trade at the same spread costs about 1 basis point.
Of course, whether a $0.05 spread is "small" depends on the share price—it's 1 basis point on a $500 stock, but 10 basis points on a $50 stock. Still, for liquid, widely held names, explicit spreads are often low enough to barely register for a low-turnover, long-term investor. Even a 0.1% round-trip cost is rarely what makes or breaks a pension fund holding an asset for a decade or more; asset allocation, management fees, taxes, rebalancing discipline, and avoiding forced sales at the bottom tend to matter far more.
That isn't to say small spreads don't add up. Across massive trading volumes, saving a few cents per share compounds into real money. And for institutional investors moving large blocks, true transaction costs go well beyond the quoted spread on a screen—they include market impact, information leakage, and overall execution quality.
Even so, it's the marginal picture that gives me pause. Turning a wide, fragmented market into a liquid, accessible one creates clear social value. But when spreads on popular stocks are already hugging a single penny—a difference most long-term savers will never feel—pushing latency from milliseconds down to microseconds often becomes more about competing for queue priority, order flow, cancellation speed, and protection against adverse selection.
Those capabilities are undeniably valuable to trading firms: they determine who gets filled, who manages risk better, and who captures the P&L. I'm just less convinced that they always create proportional downstream value for the average saver. An industry can pour immense investment into data, hardware, specialized networks, and top technical talent, yet the net effect can sometimes lean more toward redistributing trading profits among sophisticated participants than meaningfully lowering costs for everyday investors.
This is where my main personal reservation lies: high private returns don't automatically imply proportional social value. Being faster, more complex, or more profitable doesn't necessarily make an activity more meaningful.
During my time around trading desks as a "trader", it was completely natural to spend morning discussions analyzing how markets might react if a public figure, (e.g. DJT), made a particular stupid remark that day, and mapping out how to position around it. From a trading and risk-management perspective, that makes total sense—if an event can move prices, you prepare for it.
From a personal life-choice perspective, however, I realized I don't want to devote the bulk of my intellectual energy to preparing trading playbooks for unpredictable headlines, or competing for slight advantages in speed and friction within markets that are already remarkably efficient. That isn't a knock on the people who do this work, nor is it to say the industry is merely rent-seeking. It simply means that, for me, it doesn't feel like the right anchor for a long-term career.
My takeaway isn't that finance is meaningless. Its core functions remain vital: helping businesses raise capital, channeling savings into productive long-term investments, allocating risk to those best equipped to hold it, and keeping markets functioning smoothly.
At the same time, its value shouldn't be taken for granted. The question worth keeping in mind is a simple one: does a particular activity genuinely improve financing, risk-sharing, and long-term capital formation—or is it primarily a race for speed, complexity, and positioning inside an already efficient arena? And ultimately, is the difference it makes worth the talent, attention, and years of a career it asks for?
Written in late 2026, with the help of ChatGPT, Claude, and Gemini.