Friday, August 26, 2016

Misselling and more – Narcissism of large (but non-understood) differences

The most talked-of and debated element of financial services is the spectre of misselling by financial intermediaries, especially banks. Investors are agitated about it, regulators fully engaged with the issue, and above all, the popular media finds it a fertile ground for interesting coverage. Like most matters technical though, the mass media commentary is conspicuous by its lack of understanding of the basics, leading to much ado about very little on one hand, and missing the elephants on the other.

The latest in the series of "misselling scoops" is a study published by Mint, somewhat presumptuously titled 

Here is proof that banks mis-sell

For an article based on a claimed "academic survey", its a rather strange claim to make. Typically, market research surveys, whether on consumer behaviour (widely used by marketers), or indeed on elections seen frequently, are indicative. They are NEVER claimed to be definitive proofs of results/outcomes. Products often bomb in the market, election predictions often go wrong - simply because sampling is often variable, standard errors often large, and data interpretation frequently tricky. Therefore to posit a survey as definitive proof is a rather non-academic approach!

Now, on to the survey. Lets start with the research survey design. 

Survey Design
One, the survey is restricted to one city, Delhi, and excludes a large universe engaged in the Wealth Management space (foreign banks). Given that the Null Hypothesis of the survey WAS NOT "Private and Public sector banks in Delhi missell", its a rather elementary error.

Two, and somewhat less elementary, is the profile of "surveyors". The paper says that some of the surveyors were only graduates, some post-grads. And they had to be "trained on basic financial concepts, on the plethora of tax-savings products available in the market, and on how to ask for advice in the bank". Mystery shopping on technical questions are almost never done by novices. Given that questions and their answers are open to interpretation, follow-up questions require a degree of understanding on the issue at hand, these invariably need to be carried out by folks who have typically "been there done that". Surveys on airline pilot safety behaviour, for example, is never carried out by folks who have been trained for a couple of days on an X-box aircraft game! 

Now, lets see some of the key assumptions underlying the survey.

Hypothesis of banks as “financial advisors”
The study starts off with the hypothesis that banks (and bankers) need to act as advisors, assessing the financial situation of the client, do needs analysis, future projections and then recommend suitable products. These indeed, come under the fiduciary responsibility of Investment Advisors (IA). However, bulk of financial intermediaries are NOT certified as IA at all, but only as distributors, many as simply corporate agents of the tied Insurance Company! Neither distributors not Corporate Agents of Insurance are responsible for, indeed even authorised to, indulge in financial planning activities. Extant SEBI regulations are quite clear on that point. Its a bit like doing a survey in a sample of a mix bunch of general physicians, specialist Orthopaedics, nurses and hospital janitors over questions on knee replacement surgery!

The study however does no stratification of the sample to ascertain whether the bankers tested are certified as advisors, distributors or corporate agents! Most likely, a vast majority of bankers interviewed are not certified as advisors at all. A large number of (especially PSU) bankers might not even be certified as Distributors.

Ergo, it is but natural that non-certified bankers would simply point out default options (like FD) to clients looking for open-ended “tax saving” instruments. It is also understandable that bankers that are only certified as Corporate Agents of a tied Insurance company would tend to suggest insurance policies – they are massively popular as tax-saving instruments (and the popularity predates the current brouhaha over misselling).

Key issue is simple – the survey assumes that all bankers are qualified and appropriately certified to be IAs, and expects appropriate fiduciary behaviour from them. The fact is that extant regulations don’t permit such universal coverage, and for good reason too.
While the study loosely references regulations like UK RDR, it seems to have had no appreciation of differential roles and responsibilities intrinsic to the make-up of such regulations. The basic rationale and premise of the survey, hence is problematic.


Force-fit comparison of wildly divergent products
FD, ELSS, Endowment Insurance, ULIP – the study takes four wildly divergent product categories and force fits a “choice” between them. Finance 101 tells us there’s none at all. All these products lie at different points of the CAPM curve, some don’t at all – Insurance products have bundled protection elements that make their payoffs rather non-linear to fit into a CAPM curve.


Three, attribution of “costs”
The study starts off by making an astounding claim that bank FDs have “zero cost” to the investor. All banks would be out of business if that were to be true! Banks make a margin over and above the cost of deposit plus a credit charge – that’s how the bank makes money! Banking 101, missed by folks who haven’t been bankers themselves J. This gross misinterpretation is symptomatic of the general understanding of “costs” on the part of the survey.

For 3rd party products, the survey conflates the issue of intermediation costs with total costs of investment products. Investment performance, in inflation (and tax)-adjusted terms is a function of total costs in-built. Intermediation costs are a subset of total costs, and is simply a retrocession from the latter. Comparison using incomplete subsets is an obvious mistake that even a non-academic study would avoid.

The biggest bugbear of the study being Insurance, the study never discloses the total 20 year costs of an insurance policy, compared to the total 15 year costs of an ELSS (MF), or 20 year costs of FD. As a result, misses multiple issues with the formulation. First, the animal of a “20 year FD” practically does not exist. Banks do not raise ultra long tenor funding. Ergo, with differential tenor, comparing costs are a non-sequitur. Second, neither FD nor ELSS have bundled mortality covers that Insurance policies have. The authors could have (and should have) unbundled the mortality charges to arrive at comparable costs. Third, ELSS are equities-only products. Held for 15/20 years, they wouldn’t reflect the changes in risk profile that an individual undergoes over such long periods (ULIPs typically have asset allocation features to take care of this). Again, comparison therefore stands to be rather moot.


Net net, for a survey-based study that is so full of gaffes, the news report rather presumptuously claims finality and proof for its null hypothesis. Fact is, misselling is a complex issue, doesnt lend itself to simplistic formulations and "only" survey-based answers. Unfortunately, such conclusions result in erroneous conclusions, builds the wrong perceptions, and becomes a huge hindrance to market development.