LTCI Time Bandit

A time-series analyses on a set of Long Term Care Insurance (LTCI) carrier policy forms (aka Books).

A couple weeks back, I had a meeting with accounting pros to evaluate the veracity of LTCI industry’s narrative. A question came up that prompted me to perform a time-series analyses on a set of carrier policy forms (aka Books). The work I co-authored for LTCI Rate Adjudication and Neutral study (Oct, 2019) required only latest carrier filings which included rate history to understand premium rate projections. The new question had to do with trending of claim expense projections (CEP) which was not within the scope of the Oct. study.

Here is an excerpt of the never changing Executive Summary boilerplate used in CT for at least the last 10 years for nearly all rate filings:

The company says it is seeking the increase because actual claims costs far exceeded projected costs that were calculated when the product was originally priced. Unlike medical health insurance with premiums set to cover expenses incurred only during the upcoming policy year, long term care premiums are set to cover expenses that are not expected to occur until a distant date, sometimes 20 years in the future. After an actuarial review, the Department determined that the experience on this closed block of business is worse than expected and continues to deteriorate. The statutory lifetime loss ratio of 60 percent has already been met. As a result, the Department agreed that a rate increase is warranted…

Is this true? Well, only a wee bit.

Time Series Experiment

For purposes of an experiment, we used the recent PRU filing as a starter though the discussion below applies to the industry as a whole. First, we introduce terminology from our Oct. research paper that is helpful in answering the question:

    • Step Up (SUP) is a one-time increase to bring the book immediately to CT’s statutory lifetime loss ratio of 60% using LTCI industry’s claim-based model (f-CBM) which we consider flawed. It makes the present inforce responsible for past losses due to carrier under-pricing (mistakes).
    • Rate Neutral (a.k.a. Step Down, SDN) does not make the present in-force policyholders responsible for carriers’ past losses (mistakes).

The Step Down model has these benefits and objectives: (1.) provides future rate stability; (2.) eliminates discriminatory pricing for age-up policyholders; (3.) restores the concept of a level premium product as was intended & suggested at the time of sale, and (4.) provides a contractual level playing field between carriers and policyholders. The Rate Neutral model is the endpoint for rate stabilization or true ups for past overpricing resulting from excessive rate grants. 

For the experiment, we focused on a PRU’s ILTC1 rate filings (CEP 2012, 2016, 2017, 2020) limited coverage only.

Referring to the table below, the metrics of Step Up (SUP) and Down (SDN) represent a multiple of original premium (base = 1.0) now corrected for the modified CEP as actuaries change their CEP in these filings. The SDN metric is reflective of true CEP in a filing and should be considered fair pricing. SUP^ and SDN^ is the growth in both metrics relative to the base year (2012). CB% is the % chargeback if premiums were stepped up now to their SUP values. SDN^/SUP^, the relative growth ratio, is the ratio of CEP to SUP growth.  A ratio < 1.0 is a measure of the degree current policyholders are charged for past losses discounting the effect of CEP. As you can see in the table, the ratio diminishes with time, not a good thing for PHs.

Year SUP SUP^ SDN CB% SDN^ DN^/UP^
2012 4.54 2.43 46%
2016 11.25 248% 3.52 69% 145% 58%
2017 13.5 297% 3.38 75% 139% 47%
2020 19.7 434% 3.52 82% 145% 33%

Why the disparity between SUP and SDN ? While ILTC1 SDN’s moderate growth is tied to deteriorating claims experience, the SUP uses the industry’s flawed claims-based model (f-CBM). A SUP near 20 (in 2020) means premiums should be 20x original using f-CBM. Why so high compared to earlier years? (1) It is an old legacy Book well into its premium life-cycle and, (2) the remaining policyholders are asked to shoulder the burden in a narrowing time window when most premiums should have been paid by now over the Book’s life.

The SUP and SDN difference reflects a past loss chargeback if theoretically regulators were to permit a Shock Lapse (won’t happen for political reasons). Example:  Take 2017. 13.5 (SUP)- 3.38 (SDN) = 10.12, this chargeback would be a shocking 75% (10.12 / 13.5) of the premium. Ridiculous, right? 2020 and beyond would have an increasingly large SUP even if the CEP (SDN) were to remain constant.

Could CEP decline (i.e. claims projection improving!)? In performing a What If Analysis — by taking other claims projections (e.g. use 2010 in place of 2020’s), you find that the SUP / SDN ratio stays about the same and both would be lower yet the relative growth ratio (SDN^/SUP^) stays constant. Still, the carrier might find it profitable to file for a rate increase (if the minimum loss ratio remains above 60%).

The Executive Summary Question

Is their narrative true? Answer: The Executive Summary is misleading because it omits certain critical information — the effect of not citing the industry’s attempt to recover past losses to those policyholders who remain in the Book.

A final note: The CEP of most other carrier filings is often far less volatile and lower than PRU’s ILTCI1. The above discussion is very important for those Consumer Activists or their agents to consider for any carrier.

Another Over The Top In CT

With the increase, ILTC1 still remains at a stratospheric CT lifetime loss ratio of 141%. If you do the math, ILTC2 is also underpriced by over 50% and that is for a vintage sold 11/2003 – 6/2008.

This post is prompted by Prudential’s PRUD-132273692 filing, which includes policy forms ILTC1 – ILTC3. Our interest are ILTC1 and ILTC2, the older books of the filing. The requested increases are:

                • ILTC1 / +176%
                • ILTC2 / +103%
                • ILTC3 / +48.9%
                • ILTC3R / +24.5%

Noteworthy is the late start for ILTC1, the oldest book of the 4 (pre-2000), an 18% increase in 2016 followed by 2 rate increases of 10% (2019) and 15% (effective 2020).

This filing has the same ring as a recent METLIFE filing, one that contains several books across different eras or vintages. For us it is convenient that different vintages appear in one tidy filing as it allows us to more easily construct a consumer timeline & narrative of the LTCI industry.

With respect to the pre-2000 ILTC1, one might ask — what has suddenly changed to request such an increase in one gigantic gulp after twenty years? Policyholders would wind up paying 4.12x original premium if granted, but we wonder if the CT DOI, like they did with Brighthouse’s +173% Well Over The Top request, will table it to bend to increasing pressures within CT and hope for a Hail Mary from NAIC this Fall.

Pending CT legislation known as SB329 proposes the following addition:

If the commissioner determines, in the commissioner’s discretion, that an insurance company, fraternal benefit society, hospital service corporation, medical service corporation or health care center deliberately or recklessly included a misstatement of fact in, or deliberately or recklessly omitted a statement of fact from, a rate filing filed on or after January 1, 2021, that caused a long-term care policy to be underpriced by at least fifty per cent, the commissioner shall refer such rate filing to the Attorney General for an investigation pursuant to section 5 of this act.

I am on record opposing this addition which is redundant with Connecticut General Statutes 38a-665 that disallows the underpricing of insurance contracts already. The question might be which pre-2000 LTCI product isn’t underpriced by 50%? If you do the math, ILTC2 is also underpriced by over 50% and that is for a vintage sold 11/2003 – 6/2008.

What is next for ILTC1? With the increase, ILTC1 still remains at a stratospheric CT lifetime loss ratio of 141%. Our analysis shows that if priced at the statutory minimum loss ratio of 60%, the actuarially justified Shock Lapse premium would be nearly 20x what the ILTC1 policyholder paid in 2015. Makes sense, right? In lieu of the unlikely Shock Lapse scenario, ILTC1 policyholders can expect to experience a 14.5% average compound annual increase until they lapse or go on-claim, whenever than might be.

This is not to pick solely on Prudential. Here is written public testimony from an individual irate about the same issue with Transamerica. After all, the LTCI industry is the problem. You can hear the March 10th testimony including the author who submitted testimony somewhere in the middle.

The SB329 excerpt above seems written more for the industry than the consumer. We track LTCI legislation and will comment further as the legislative session proceeds. Another concern we see in this bill is the emphasis on affordable benefit options (ABO). What does affordable convey? Why the need to put into legislation? Of course we know the answer but will cover later if it stays in legislation. The whole point with SB329 is that it seems to be especially targeted to dismissing early legacy policies, such as ILTC1 and ILTC2.

Update: Examining PRU filings since 2016, we have noted an important change in assumption of their morbidity model that increases claims expense projections and lifetime loss ratios, which are detrimental to policyholders. Against the 2020 filing, we employed What If analysis to arrive at an increase of 0.26x (8%) from 3.25x to 3.51x original premium using our Step Down model (considered fair premium) resulting from the changed assumption. One might consider 8% noise compared to +173% requested increase, but we wonder what other subtle changes might be going on below the surface? Note: Step Down model does not hold policyholders responsible for carrier past losses.

 

 

%d bloggers like this: