Source – Wikipedia
The physical DSA’s (Direct Sales Associates) as a breed is disappearing fast and the said is being replaced by Digital DSA’s (Part of Fintech revolution). I call them Digital DSA’s because they would still need to do many tasks which the traditional DSA’s do when compared to only a referral channel (Pure Lead Generator).
The latter ones are being setup by Tech Guys, ex-Bankers etc. Many of them are raising money or have raised money from Investors.
They have many advantages over the traditional DSA’s such as Wider Reach, Being able to facilitate credit from more than one lender which the traditional DSA could not on account of being tied up to an exclusivity contract. The said obviously allows them to have a higher volume and a higher through put of leads to converts. This coupled with their claim to add value in terms of augmenting the credit application with additional data points and lastly a smart tech based learning which based on past history and additional data points will allow the Digital DSA’s to eventually match leads to the right lender thus ensuring a faster TAT for them & their customers along with throughput.
All of these are a clear edge over the traditional DSA’s. If I were to look at the evolution and subsequent decline of physical DSA’s, I see that High DSA commission due to lack of reach, Higher ROI to borrowers & Productivity of FOS were the primary determinants of their success with Lower Commission on account of Lower ROI’s to borrower, Physical Bank Reach and Tapering Productivity leading to their decline.
In case of Digital DSA’s it would be wrong to assume that Lenders would not change their commission structures & further pass on benefits inform of lower ROI as there will be multiple Digital DSA’s and this coupled with fact that lenders are beefing their own Digital Channels thereby offering seamless credit to existing customers as well as prospective borrowers.
As on date many lenders have special ROI’s on Personal Loans, Auto Loans, Housing Loans for their captive customers holding CASA accounts or existing lending relationships.
A borrower will go to his/her lender/bank separately through their digital channels along with simultaneous inquiry at a Digital DSA. In such a case the throughput of the Digital DSA will get impacted. The Digital Channel of the bank on account of seamless view they hold of their customer will be able to offer a Faster TAT and Lower ROI. The probability of such customer not finally availing disbursal through digital DSA is very high, clearly a case of lower productivity for Digital DSA.
Having lend, I know that throughput (Files Logged to Disbursed) in the private/mnc lender space (banks & nbfc’s) does vary but is not more than 1.5x between the lowest throughput lender and the highest throughput lender for a particular target market and specific lending product. Also a point to be noted is that higher throughput is also linked to higher rates for the borrower in majority cases.
As on date CIBIL has over 350 million individual credit records and around 20 million in the Non retail space. This is 10 times of what existed about a decade ago when additional data points would have helped doubled or quadrupled the throughout (Files Logged to Disbursed) for a DSA. It is a matter of time before Utility Records become a part of CIBIL score. If we are talking about 60% of population living below USD2 per day, CIBIL practically will cover 80-90% of credit prospects from remainder of the population.
Coming to the question of whether additional data points over and above past credit history will enable Digital DSA’s to have a higher throughput (Leads to Converts/Disbursals), the answer is not that simple.
Internal lending parameters augmented with CIBIL scores have shown that Lenders have been able to lend rightly. Rightly would mean lending to a borrower who has the ability to repay with a fairly accurate inference on intent to repay (This one is always tricky) and ability to recover/repossess.
The credibility of newer data points to make a case for existing lenders to increase throughput is too thin. Lenders are in business of lending & not distributing money. There is no evidence to make case out for alternative data points as bridging or providing cushion for shortfall of 100-200 points in CIBIL score. In fact many lenders have run & do run surrogate programs which are akin to making lending decision on alternative data points which augment current lending parameters. This surrogate programs are based on a very large number of alternative data points available within their own system and yet these programs are not the mainstay of lending criteria and throughput is definitely not double in case of such programs. Infact many such surrogate programs in the last decade were a pain point for many lenders.
That brings me to the third point of machine learning capability at Digital DSA’s of matching leads to borrowers based on past throughput. It is going to be a tradeoff between commissions & throughput for them.
Machine learning would be more effective for non-retail borrowers seeking working capital, business credit facilities wherein the availability of past credit history is still an issue and assessment of repayment capability varies from lender to lender. Non retail including SME’s have been a riskier segment for most lenders and hence ability to match borrower to lender would work as long as credit policies are stable. The productivity of staff & Attrition at Digital DSA’s would make or break their profitability when catering to this segment.
There is an interesting case for Digital Lenders who have raised money wanting to use alternative data points for catering to Non Retail space. As a lender the tradeoff is between better credit (quality) rather than deeper credit (Quantity). This tradeoff is always predisposed towards Quality and hence to presume that such lenders will provide 2 or 3 times higher throughput as compared to current lenders does not seem plausible inspite of them raising money with premise of trying to repair or provide an alternative lending system.
Majority in the Credit Risk assessment space very well know that any reasonable effect positive or negative of change in a lending assessment starts showing in lending books after 15-18 months. I doubt if the new Digital Lenders would be that brave to augment 200 points on CIBIL score with alternative data points which they curate directly or through Digital DSA’s or wherein credit history is not available be that brave to offer twice the throughput as compared to the incumbents.
The picture is not that rosy as is made out by the media. The size of commission/fees market as on date for retail segment in Personal Loans is around INR 250 cr of which not more than 30% or 40 % would be available for Digital DSA’s. The Auto loans segment would be INR 1100 cr. Of this not more than 30% is available for Digital DSA’s as POS channels (Dealers) are pretty strong & their ability to reverse subvent loan commissions to get Car Sales cannot be matched by Digital DSA’s. Why would i not want extra accessories as a borrower/car owner for free if it comes as part of availing loan through Dealer DSA and that too without compromising on ROI. For Housing Loan again the POS (Builder Sales Desk) works as an effective lead generator feeding into Direct Sales Force of Lenders and Physical DSA’s as well.
The game for Digital DSA’s is very clearly driving volumes through their channels by maintaining a lower Cost of Acquisition. They should not forget about the learning’s from declining breed of Physical DSA’s, must not underestimate current lending configuration (Internal Lending Parameters plus CIBIL) which in this decade has worked very well for Lenders. Lastly they must not forget that prospective borrower will go to more than one Digital DSA at any given time & in parallel also approach his existing Bank/Lenders Digital Channel.
There is a fairly large underpenetrated market of those earning below USD 2 per day but that is a segment which is being addressed by the MFI’s to some extent, wherein ticket sizes are small & risk is higher. The latter is not the segment amenable to Digital DSA’s and unlikely a business case even if they want to address. This segment will eventually transit but that is for the future and dependent on how we eventually perform as an economy in this decade and coming decade.
Digital DSA & Digital Channels of Lenders will bring about a paperless process as is detailed below . The credit evaluation at banks end will get automated through parametric approach based on repayment capability, credit scores & authentication of data points for many lending products. The Branch Credit Manager may have no role to play other than for discretionary cases. Move towards paperless processing has started for simple products like PL wherein it is on the TAP in case of many banks for existing customers. This however does not fundamentally change the throughput (leads to converts) but will help prevent fraud. Automation resulting in Faster TAT could also mean lower commissions for Digital DSA’s. Such a scenario in near future would also mean shakeout within the current Digital DSA’s. The Digital Channel of Lenders will also get available to Car Dealers & Builders as part of maintaining influence over the POS channels.
1) Borrower selects his requirement (PL/Auto/Housing Loan) through Digital DSA or Digital Channel of a Lender
2) Enters Pan card/Adhar Number
3) Application form auto inked from past data or drawn from third party sources
4) Application is digitally signed by borrower
5) Auto KYC or KYC Enabled through Adhar Authentication or Utility Company or MCA for non retail gets in motion
6) Proof of income validated through existing Salary Account or ITD/MCA in case of Professional/Non Retail
7) Auto CIBIL Validation
8) Real Time Loan Approval/Rejection
9) Borrower acceptance of terms of lending
10) E-Lien creation with Mortgage Repository or RTO in case of Housing/Auto Loans after acceptance by lender
11) Disbursal in Wallet/Bank account/Builder Account/Dealer Account
The original article was published on linkedin.
@kalpesh_s_desai
