Understanding Commissions Motivated Advice: Evidence from Indian Life Insurance Santosh Anagol (Wharton), Shawn Cole (Harvard Business School), Shayak Sarkar (Harvard) Wharton, February, 2012
Introduction Many financial products difficult to value, particularly for those with limited financial experience (mortgages, life insurance) Little learning for long-horizon products May limit usefulness of brokers to build reputations for providing the right products Research Agenda: How do consumers make decisions about these complicated financial products? Research Questions Today: What is the quality of advice that commissions motivated agents provide? Under what conditions does advice improve? Many other inputs into consumers decisions: Press? Friends? Regulations? Government campaigns?
Motivation: Why Life Insurance in India? Why India? Increasing incomes in China, India other fast developing countries will greatly increase capacity to invest in formal financial products How will these consumers make informed decisions? What role should government play? Important question in the U.S. as well (creation of Consumer Financial Protection Bureau) Why Life Insurance? 20 % of Indian household financial savings in life insurance products Easiest product to identify potentially bad decisions Approximately 2.4 million life insurance agents in India (approx 434,000 in the US)
Focus on Term vs. Whole Most popular products Easy for us to compare/evaluate Term : pay P T for N years, receive payout C T at death during that period, or nothing if survive N years. Whole: pay P w per year for N years, receive P w N + B at min (year of death, max(40 years after purchase, age 80)) Surrender value: 30% of premiums paid if paid > 3 years How are bonuses (B) determined? Discretion of life insurance company A percentage (typically 3%-5%) of sum assured (P w N) Importantly, not compounded Whole type products have estimated 60-80% market share
Replicating Portfolio Consider Rs. 500,000 (ca. $10,000) coverage for 34 year-old male Whole life policy costs Rs. 13,574 per year, paying 3% bonus Term policy for equivalent coverage (Rs. 500,000) and save remainder 2,507 per year + 11,067 savings deposit (earning 8%) for 25 years (until 2035) Savings contribution 13,574 from 2035 until 2056 Clear violation of law of one price If you die before 2056: almost surely better off with term + savings (savings are liquid) If you survive until 2056 Whole redemption value: Rs. 1,205,000 Savings balance: Rs. 5,563,378 Note: no risk of future premium increases for term product (Cochrane (1995)) Rajagopalan (2010) has similar findings
Why Would Anyone Choose Whole? Agent receives commission of 35% on whole, 5% on term Paper presents model where a dominated product with high commissions can exist in competitive equilibrium Buyer cannot calculate effective cost Term is throwing money away if you survive until the end of the policy, it s worth nothing People don t appreciate importance of compounding (Zinman and Stango (2009)) Whole policies pay 3%-5% bonus per year not compounded! Commitment to save Why does commitment to save have to bundled with insurance? Public provident fund is a commitment savings product paying compound better returns
Audit Study Hire 10 auditors, making a total of 1,026 visits to insurance agents over 12 months Field experiment conducted in two major cities in India Audit process developed by a former life insurance salesman from major bank Agents found on publicly available yellow pages type websites Week-long training, practice audits Each auditor has personalized (true) script ( I am a married man with two kids... ) Certain features disguised My salary? Let s say I earn Rs. 10,000 per month
Channels of Life Insurance Sales Distribution Channel (1) Individual Agents 79.6 Banks 10.6 Other Corporate Agents 4.30 Brokers 1.38 Direct Selling 4.13 Source: IRDA Annual Report, 2009-2010.
Pilot Script Introduce self, express need for life insurance Not looking for investment product Seeking maximum risk cover at minimum cost If need to save, prefer to save in a bank What policy do you recommend?
Pilot Script: Proportion of Term Recommendations Recommendation Risk Coverage Script (1) Only Term Policy Recommended.09 Any Term Policy Recommended.16 Only Whole Type Policies Recommended.31 Any Whole Type Policy Recommended.64 Any Other Policy Type.18 Observations 60
Agents Talk Down Term Insurance You want term: Are you planning on killing yourself? Term is throwing money away Term is not for: Women Middle class Term is only for: businessmen government employees Offered endowment policy, calling it a term policy Only one instance of explicit debiasing Don t buy whole, it s a rip-off
Quality of Advice: Multiple Recommendations Most term recommendations come as a part of multiple recommendations (a package)
Quality of Advice: Suitability and Catering Do agents provide advice based on client s actual need, or client s beliefs about what is the right product? Important question in context where clients are unlikely to understand differences in products Vary need: Whole: I want to save and invest money for the future, and I also want to make sure my wife and children will be taken care of if I die. I do not have the discipline to save on my own Term: I am worried that if I die early, my wife and kids will not be able to live comfortably or meet our financial obligations. I want to cover that risk at an affordable cost. Vary beliefs: I have heard that whole insurance is a really good product. I think it may be suitable for me. Maybe we can explore that further? I have heard that term insurance is a really good product. I think it may be suitable for me. Maybe we can explore that further?
Quality of Advice: Responding to Needs & Beliefs Overall low rate of recommending term insurance - even when auditor says they want risk coverage and have heard term is a good product But needing risk coverage does cause about 12% higher probability of receiving term recommendation Even when customer initially believes whole may be better for them At least some agents know that term insurance is better for risk coverage
Catering vs. Quality Advice: Term Insurance Dependent Variable Any Term Only Term (1) (2) Belief Term 0.10*** 0.02* [0.03] [0.02] Need Term 0.12*** 0.016 [0.04] [0.01] Belief Term * Need Term.024.052* [.059] [.031] Government Underwriter -0.12*** -.01 [.041] [.02] Constant -0.06-0.01 [0.05] [0.01] Auditor FE YES YES Observations 511 511 Adjusted R-squared 0.10 0.034 Mean of Dep Var 0.13 0.03 Agents do cater advice to both customer preferences and need for risk coverage Not just whole recommending machines But catering mainly by adding on a term policy on top of a whole policy Following the path of least resistance Government underwriters less likely to mention term plans overall
Wide Range of Risk Coverages Recommended Belief & Need = Term Risk Cover (U.S.D) Premium (U.S.D) Whole Life Type Policies 12,997 629 Term Type Policies 44,494 619 Only 10% of auditors get a term recommendation But the amount of risk coverage they get recommended is approx 4 times larger Possible theory: agents cater to premium amount that can be paid
Improving Advice: Natural Experiment on Effect of Disclosure Natural experiment on ULIP disclosures Prior to July 1, 2010, agents required to inform buyers about total ULIP costs/charges As of July 1, 2010, agents are required to provide separate breakdown of commission costs Allows us to isolate disclosure of agency problems Measure agent reaction
Field Experiment Overlaid on Natural Experiment Overlay with field experiment Agent expresses knowledge of agency problems Can you give me more information about the commission charges I ll be paying? Agent does not express knowledge of agency problems [No statement]
Results Table: 8-Effect of Disclosure on Product Recommendations Dep Var = Ulip Recommended (1) (2) Post Disclosure Regulation -0.22*** -0.21*** [0.05] [0.08] Disclosure Knowledge -0.01-0.004 [0.05] [0.07] Agent Home -0.06-0.06 [0.11] [0.11] Auditor Home -0.13-0.13 [0.17] [0.17] Agent Office -0.05-0.05 [0.10] [0.10] Auditor Office -0.04-0.04 [0.20] [0.20] LIC -0.44*** -0.44*** [0.05] [0.05] Post Disclosure Regulation * Disclosure Knowledge -0.02 [0.10] Observations 258 258 R-squared 0.35 0.35
Disclosure Results
Improving Advice: Competition? Competition and bad advice: does the threat of losing a sale to another agent make an agent more likely to match needs of customer? Vary level of competition by varying source of beliefs: I have heard from a friend that whole (life)... I have heard from another agent from whom I am considering purchasing... Does agent try to win business by correcting another agent?
Does Competition Matter for Type of Advice? Agents de-bias when advice comes from another agent Statistically significant at 5 percent level Note that this de-biasing is mainly through recommending term in addition to whole
Improving Advice: Sophisticated vs. Un-Sophisticated Customers High level of sophistication: In the past, I have spent time shopping for the policies, and am perhaps surprisingly somewhat familiar with the different types of policies: ULIPs, term, whole life insurance. However, I am less familiar with the specific policies that your firm offers, so I was hoping you can walk me through them and recommend a policy specific for my situation. Low level of sophistication: I am aware that Life Insurance products are complex, and I don t understand them very much. However I am interested in learning, what type of policy may be right for me? Delivered in introduction of auditor to agent Remainder of script unchanged In particular, stated income held constant
Sophistication Results: Product Recommendation (1) (2) (3) (4) Dependent Variable: Any Term Only Term Ln(Coverage) Ln(Premium) Sophisticated 0.10* 0.10 * 0.21* -0.06 [0.06] [0.06] [0.12] [0.10] Government Underwriter -0.08-0.09-0.25 0.05 [0.07] [0.06] [0.16] [0.10] Auditor FE YES YES YES YES Audit Location FE YES YES YES YES Observations 217 217 209 209 Sophisticated agents 10 percentage points less likely to receive recommendation of any term Sophisticated agents also recommend to buy more coverage, but not to pay more premiums - consistent with catering to premium amount story
Conclusion Quality of Advice Agents mainly recommend whole despite fact that term + savings seems to dominate Even to customers who mainly want risk coverage Agents will cater to incorrect beliefs When agents do recommend term, they prefer to do it as a package (whole + term) Improving Advice When agents forced to disclose information changes advice Some evidence that agents will compete by providing different advice When consumer signals sophistication gets weakly better advice