Executive take
Wall Street and big tech CEOs are again posting eye‑catching paydays, but the bigger question for investors isn’t the headline number—it’s the wiring underneath. In 2006, finance-industry compensation frequently rewarded balance‑sheet expansion and near‑term revenue with limited downside for mistake‑driven losses. Research after the crash argued that this “heads I win, tails you lose” payoff profile could amplify risk-taking.
Today’s pay mechanics are different in important ways—more equity, more multi‑year vesting, more formal clawbacks and disclosure—and banks operate under tighter post‑crisis capital and governance constraints. Yet equity-heavy packages still tether executives to short‑term share price and “financial engineering” channels (including buybacks), and the pay ratio data show how sharply the distributional gap can widen even in strong years.
Bottom line: Current pay practices can reinforce short‑termism, particularly when performance targets are dominated by stock price, EPS optics, or annual revenue. But the set of conditions that turned 2006 compensation into a systemic accelerant—extreme leverage, thin capital buffers, and mispriced tail risk—looks less present inside regulated banks than it did pre‑2008. The bigger “rhyming risk” may be migrating toward places where leverage and opacity are higher and red‑tape thinner (parts of market-based finance), even as CEO pay trends continue to reward equity rallies.
What 2006 “paid for” and why it mattered
A useful way to remember 2006 is that many large financial firms were effectively running businesses where balance sheet scale and short-cycle profit were prized—and where leverage quietly did the heavy lifting. The New York Fed’s work on dealer balance sheets highlighted how leverage and liquidity conditions can move together, creating feedback loops that deepen stress when asset prices fall.
Post‑crisis compensation critiques didn’t claim pay was the cause of 2008; rather, they argued pay helped set the incentives: executives could monetize upside (bonuses, equity) while much of the downside arrived later and was borne by shareholders, creditors, and (in a panic) the public backstop. Bebchuk and Spamann’s analysis is emblematic: when rewards are convex and penalties limited, decision-makers can rationally prefer riskier strategies even if they reduce long-run firm value.
Academic evidence also complicated the popular narrative that “bad incentives” simply meant CEOs weren’t aligned with shareholders. Fahlenbrach and Stulz found that banks where CEOs were more aligned with shareholder interests did not fare better during the crisis—and in some tests did worse—consistent with the idea that shareholders themselves may favor risk when losses can be shifted to others in tail events.
What CEOs get paid now — and what it’s tied to
A snapshot from proxy statements shows why pay draws attention: totals can be enormous, but the mix and measurement have changed. Large banks still deliver much of pay in deferred equity, while big tech increasingly runs on stock awards and performance metrics; and modern proxy rules also force companies to disclose “pay versus performance” tables, including “compensation actually paid,” which can swing dramatically with stock price changes.
Representative compensation and pay ratios
| Company | “2006-era” CEO pay (reported) | Latest CEO pay (reported) | Median employee pay | Pay ratio (CEO:median) |
|---|---|---|---|---|
| JPMorgan Chase | Jamie Dimon: $39.053M (2006) | Dimon: $39.004M (2024) | $153,265 (2024) | 254:1 |
| Goldman Sachs | Blankfein tally: $54.0M (2006) | Solomon: $31.290M (2024) | $162,703 (2024) | 192:1 |
| Morgan Stanley | Mack: $41.411M (2006) | Pick: $24.881M (2024) | $138,509 (2024) | 180:1 |
| Microsoft | Ballmer: ~$0.976M (FY2006) | Nadella: $96.497M (FY2025) | $200,972 (FY2025) | 480:1 |
| Alphabet | — | Pichai: $10.725M (2024) | $331,894 (2024) | 32:1 |
Two observations jump out:
First, the distributional gap is explicit. Microsoft’s disclosed pay ratio (480:1) shows how quickly CEO pay can detach from the middle of the workforce even when median employee compensation is six figures.
Second, “reported pay” can be misleadingly smooth compared with the economics. Goldman’s pay-versus-performance disclosure shows that “compensation actually paid” for its CEO can spike far above the Summary Compensation Table total (a reflection of equity valuation changes baked into the SEC’s required methodology).
The key risk channel today: equity incentives plus buybacks
If 2006’s hallmark was paying handsomely for leverage-fueled growth, 2020–2026’s hallmark is paying handsomely for equity outcomes. That’s not necessarily bad—shareholders benefit when stock prices rise—but it can bias management toward strategies that lift near‑term share value even if they raise fragility.
One reason is buybacks: they are a legal, mainstream capital return tool, but they can also serve as an EPS and share-price support mechanism—particularly when combined with performance metrics that reward per‑share outcomes. Federal Reserve data show corporate repurchases have been a persistent, large feature of the capital cycle, while net issuance has often been negative over long stretches.
The “Bloomberg-style” point here is not that buybacks are inherently destabilizing. It’s that when CEOs are paid primarily in stock and evaluated on per‑share metrics, capital allocation can tilt toward predictable, finance-driven boosts rather than longer-dated investment—or toward cutting risk controls and headcount in ways that look good in the next quarter and age poorly in the next cycle.
That same structure shows up directly in mandated “pay versus performance” tables. Microsoft’s CEO pay illustrates the difference between grant-date totals and SEC-defined “compensation actually paid,” which varies with the valuation of equity awards.
So… does this look like 2006?
It rhymes in incentives, not in plumbing.
The rhyme is the continued dominance of convex payoff structures: executives capture meaningful upside from expansion, equity rallies, and measured “performance,” while some downside is delayed, diluted, or shared. The literature warning about that convexity—especially in banking—remains relevant.
The difference is the post‑crisis guardrails and disclosure regime. Since 2008, regulators and standard-setters have pushed for risk-sensitive pay practices, and global bodies like the Financial Stability Board published principles explicitly aimed at reducing incentives for imprudent risk-taking. In the U.S., clawback requirements and listing standards have also been strengthened in recent years, formalizing mechanisms for recouping certain compensation after accounting restatements.
Still, the Goldman proxy itself is a reminder that even “risk-balanced” systems can evolve competitive add-ons (including retention-style equity and carried-interest-like elements), potentially reintroducing complexity and new risk incentives if not tightly governed.
Practical guardrails that reduce short-termism without banning big pay
This is the governance playbook that most directly targets systemic concerns (rather than simply expressing outrage at high numbers):
Longer holding periods and slower selling windows for top executives. The risk isn’t stock compensation; it’s early monetization. The longer equity must be held after vesting, the more it behaves like true long-term capital.
Risk-adjusted performance metrics for banks and financial intermediaries. Pay should explicitly incorporate constraints like capital, liquidity, and risk-adjusted profitability rather than raw revenue or RoE targets that can be “bought” with balance-sheet risk. The rationale follows directly from the post‑crisis compensation literature and the Fed’s leverage/liquidity feedback work.
Clawbacks that are usable, not symbolic. The SEC’s reset has improved baseline standards, but boards still control how aggressively they enforce policies in gray areas.
Buyback governance tied to investment and resilience. Boards can require that repurchases do not impair stress-test resilience, liquidity needs, or critical long-run investment—especially when executive scorecards include per-share metrics that buybacks mechanically improve. Fed repurchase/issuance data underscore why this lever matters at scale.
And so….
If you’re looking for a clean “CEO pay today = 2006 all over again” headline, the evidence doesn’t support it—at least not inside the most regulated parts of the financial system. The more accurate assessment is that today’s outsized pay is less about incentivizing leverage directly and more about incentivizing equity outcomes, which can still produce short‑term behavior. And in finance, where leverage always lurks somewhere in the ecosystem, incentives that reward upside without fully pricing downside remain a perennial risk factor—one the post‑crisis literature explicitly flags.
