Editor's note. This is the closing chapter of the volume. Every preceding chapter has documented a failure — of due diligence, of audit, of inspection, of imagination. The paper reviewed here measures what those failures cost in a currency the rest of the book has only gestured toward: the durable, regionally distributed willingness of ordinary Americans to hand their savings to a stranger. That willingness, the authors show, did not recover within the window of their data. The question with which the book ends is whether it can be restored at all.
Most of the scholarship on Bernard L. Madoff asks who failed to catch him. Umit G. Gurun, Noah Stoffman, and Scott E. Yonker, in Trust Busting: The Effect of Fraud on Investor Behavior, 31 Rev. Fin. Stud. 1341 (2018), ask the question that begins where those failures end. Once a fraud of this magnitude becomes public, how does it change the behavior of investors who were never victims? Their answer, causally identified across thousands of zip codes and hundreds of advisory firms, is that the damage spreads well beyond the perimeter of the original wrongdoing. Clean advisers — firms with no connection to Bernard L. Madoff Investment Securities LLC ("BLMIS"), no exposure to the feeder funds, no whiff of impropriety on Form ADV — lost assets and exited the industry at significantly elevated rates in communities where Madoff's victims happened to live. Trust, Gurun and his coauthors show, is not a soft residual variable. It is a measurable, priced input into the demand for advisory services, and when the input is shocked, capital reallocates, intermediaries close their doors, and the bill falls on parties who never met the man who set off the cascade.
The identification strategy
The methodological move that makes the paper persuasive is its use of geography and ethnicity as proxies for differential exposure to the revelation. Extending the affinity-network mapping developed earlier in this volume (see [owens-shores-informal-networks]), the authors construct a community-level exposure measure from two variables: proximity to known direct investors, drawn from the customer list filed in the SIPA proceeding, and the local concentration of Jewish residents, the demographic group most heavily reached by the country-club and synagogue marketing through which the fraud propagated.
The dependent variable is changes in assets under management at registered investment advisers ("RIAs"), reported annually on Form ADV under the Investment Advisers Act of 1940. Because RIAs must disclose the geographic distribution of their clients, the authors can build a panel that tracks, firm by firm and year by year, whether post-December 2008 withdrawals concentrate in advisers whose clients live in heavily exposed communities. The corresponding inflow side is captured from the FDIC's Summary of Deposits.
The results are stark. In high-exposure zip codes, AUM at advisers fell sharply relative to control communities in the years following the fraud, while bank deposits rose by an offsetting amount. The shift is not explained by local wealth shocks, by the broader 2008 market collapse, or by the presence of Madoff investors themselves — the authors strip out direct victims from the estimation. What remains is the behavioral response of uninvolved investors who lived near, worshipped with, or socialized among people who had been defrauded.
Advisers as collateral damage
The most striking finding for a law student is what happened to the advisers themselves. In heavily exposed communities, advisory firms with no connection whatsoever to Madoff — firms running ordinary long-only equity portfolios, charging conventional fees, with clean Form ADV disclosures — were significantly more likely to deregister and close. Gurun, Stoffman, and Yonker estimate that exit rates in the most exposed quartile of zip codes were roughly half again as high as in the least exposed quartile.
The doctrinal valence is worth pausing on. Section 206 of the Investment Advisers Act, 15 U.S.C. § 80b-6, imposes a federal fiduciary duty whose anti-fraud enforcement — like enforcement under Securities Exchange Act § 10(b) and Rule 10b-5, 17 C.F.R. § 240.10b-5 — is premised on the idea that the harm of fraud is harm to the defrauded client. The empirical reality the paper documents is broader: a single fraud at one adviser inflicts a negative externality on every other adviser whose clients share a community with the victims. The injury runs through the trust network rather than through any contractual or fiduciary relationship, and existing private rights of action do not reach it.
The behavioral premium on trust
The paper's most theoretically interesting result concerns which advisers weathered the shock. Firms whose business model emphasized trust-building practices — in-person meetings, longer average client tenure, smaller AUM-per-client ratios — experienced significantly smaller withdrawals than industrial, arm's-length competitors in the same exposed communities. This is the cleanest empirical evidence in the finance literature for what behavioral economists since Luigi Guiso, Paola Sapienza, and Luigi Zingales, Trusting the Stock Market, 63 J. Fin. 2557 (2008), have called the trust premium: investors will pay, in foregone returns and higher fees, for relationships that feel personal. It is also the price tag on the theoretical account of trust mapped out earlier in this volume (see [stolowy-trustworthy-investment-opportunity]).
The result cuts two ways. It vindicates the Commission's longstanding emphasis on Form ADV Part 2's plain-English relationship disclosures, and it puts numbers behind the institutional argument for stronger gatekeeping that Rhee develops in the regulatory-failure chapter (see [rhee-madoff-market-regulatory-failure]) — Gurun and his coauthors supply the price tag the gatekeeper-failure literature has lacked. It also implies that the most trust-intensive business models, precisely because they generate the deepest loyalty, may be the most vulnerable to abuse. BLMIS scored extraordinarily high on every trust proxy the authors measure. The behavioral premium, in other words, is also the affinity-fraud premium.
What restoration looks like
If trust is contagious, what does restoration require? The post-Madoff regulatory record gives a partial answer. The Commission's 2009 amendments to the custody rule, 17 C.F.R. § 275.206(4)-2, imposed surprise annual examinations by an independent public accountant for advisers with custody of client assets, and added Form ADV-E as the cover sheet for that accountant's certificate. The Dodd-Frank whistleblower regime, codified at Securities Exchange Act § 21F, 15 U.S.C. § 78u-6, provides bounties of up to 30 percent of monetary sanctions over $1 million — a structural answer to the Harry Markopolos problem (see [markopolos-no-one-would-listen]). The PCAOB's interim broker-dealer auditor inspection program, authorized by Dodd-Frank § 982, finally subjected the auditors of broker-dealers to the same inspection regime that has applied to issuer auditors since Sarbanes-Oxley. Each of these reforms is real, and each addresses a specific institutional failure cataloged elsewhere in this book.
None of them addresses the externality Gurun, Stoffman, and Yonker identify. Surprise audits restore confidence in custody. Whistleblower bounties recruit informants. PCAOB inspection raises audit quality. None of these levers reaches the widow in West Palm Beach who moves her portfolio from a clean adviser to a savings account at less than 1 percent because her neighbor's net worth has evaporated. The trust she withdraws was not regulated into existence and cannot be regulated back.
Doctrinal takeaway
For the law student who will spend a career advising registered advisers, broker-dealers, and their auditors, the regulatory architecture that emerged from this fraud is the operative landscape: the amended custody rule and Form ADV-E, 17 C.F.R. § 275.206(4)-2; the Dodd-Frank whistleblower regime, 15 U.S.C. § 78u-6; the PCAOB broker-dealer auditor inspection program under Dodd-Frank § 982; and the SIPA net-equity regime as construed in In re BLMIS, 654 F.3d 229 (2d Cir. 2011). Each of these is a mechanism designed to catch the next BLMIS. The harder question, and the one this book ends on, is whether financial regulation can restore the social capital that fraud destroys — or whether the durable consequence of a Madoff is a quieter, regionally distributed retreat that no statute reaches. The opening chapter of this volume showed that the SEC's institutional failure is, in principle, reformable (see [sec-oig-509-madoff-investigation]). The closing chapter is more sobering: the regulatory record is reformable; the trust it shattered may not be.