Some thoughts on Life Insurance:
These are simplistic views, possibly full of errors. Please point them out -
(thread)
These are simplistic views, possibly full of errors. Please point them out -
(thread)
Insurance penetration at about 2%-3% is touted as the reason for growth potential of the sector. According to 2012 McKinsey study 12% of all deaths in India are insured. This implies about 50% of all households are insured. These numbers must have gone up since then.
2/n
2/n
The percentage of population that can afford insurance has probably already been tapped. But most insurance in India is sold as a savings product and not a protection product.
3/n
3/n
Term insurance (probably the best financial innovation in the world ever) has a low penetration. This is due to legacy issues, commission structures, profitability, capital requirements of various insurance products.
4/n
4/n
ULIP is primarily like a Mutual Fund with protection component being 5% to 10%. Participating Policies are again savings products where the returns are indicative (not assured). Non Participating Policies have assured return components.
5/n
5/n
ULIP and PAR policies are essentially pass-through products. The Insurer is essentially like an AMC managing the assets for the policyholder. Non-PAR policies need active hedging there is a lot of debate as to whether these risks are easy to hedge.
6/n
6/n
At least one large insurer has been extremely cautious on Non Par policies. Interest rate hedging is the main concern here and hence (probably) carries higher balance sheet risks.
7/n
7/n
We don’t have a long history of listed insurance players and their accounting and valuation is more complex than even Banks. But for a simplistic view I compare some of the parameter to those of a Bank (this is not technically correct but helps me have a broad feel).
8/n
8/n
Embedded Value: This is broadly sum of capital plus value of all policies in force to the insurer. Actuaries (a highly evolved species that eats Bayes Theorem for lunch and gets orgasmic thinking about multivariate regression) determine this.
9/n
9/n
Lot of jazz goes into its calculation (views on equities, interest rates, mortality). In case of a bank the calculation of NPV is relatively simpler. But I would trust the EV of an average insurer more than the book value of an average bank.
10/n
10/n
Value of new Business (VNB) is the value of new policies (expected eventual profit to the insurer) for the new policies issues. People focus on APE (Annual Premium Equivalent) as it is released monthly but VNB is what matters I think.
11/n
11/n
As ULIP has low protection component, fall in ULIP sales might result in low APE but a smaller rise in Term Insurance sales might make up for the VNB.
12/n
12/n
EV for current year is EV for last year + VNB + (some other stuff) + variance (change in EV due change in assumptions that have gone into EV calculation like interest rates or persistency – how many policy holders will continue paying their premiums).
13/n
13/n
EV (Embedded Value) is like the Book Value for a bank. ROEV is analogous to ROE and VNB to PPOP (operating profit) of a Bank. In valuation terms P/EV for an insurer is like P/B for Banks (again, this is quite simplistic).
14/n
14/n
The float (premium collected) is invested on behalf of policyholders in case of ULIPs and PAR policies. The insurer can also make investment with these on behalf of shareholders. Large Pvt players for example have between 4%-23% of asset in equity on behalf of shareholders.
15/n
15/n
Similarly they do have investment in government and corporate bonds. Change in equity/bond prices have a bearing on both EV and VNB.
16/n
16/n
The best (subjective) insurers in India are trading at P/EV multiples similar to P/B multiples of best banks. The key questions are:
17/n
17/n
Is the threat of book value impairment for banks higher or lower than threat of EV impairment of insurers?
Is the VNB growth going to be lower/higher than credit growth of banks?
Is the ROEV going to be lower/higher than ROE of banks?
18/n
Is the VNB growth going to be lower/higher than credit growth of banks?
Is the ROEV going to be lower/higher than ROE of banks?
18/n
Why this comparison to banks? Because most of us understand that sector relatively better and to me at least gives a benchmark.
Pandemics generally have been positive for Insurance growth.
19/n
Pandemics generally have been positive for Insurance growth.
19/n
Why do Indian Insurers trade at such high multiples compared to regional peers? Well, why do our banks?
Also we need to value insurers as AMC + Insurance SOTP model probably.
Balance sheet risks, solvency ratios and profitability are all quite different.
20/n
Also we need to value insurers as AMC + Insurance SOTP model probably.
Balance sheet risks, solvency ratios and profitability are all quite different.
20/n
Now if you feel I or you understand insurance:
21/21
21/21
