Value Life Corp announces a new and innovative application for its GPS software (the "Global
Pricing System for Life Insurance Policy Valuation™"). GPS includes several
features to give users a
power-assist in performing quality
assurance reviews. These features are useful when using GPS as a primary tool for
pricing life settlements and other valuations of life insurance policies. When an
overworked pricing analyst may see as many as thirty to forty cases per day, even highly
qualified and disciplined analyst staff need all the information and automated assistance
modern technology brings right to their desk. Moreover, the Quality Assurance Program
("Q&A™") serves as a secondary tool for firms that use some other software as their
primary pricing tool. In any scenario, you can use your GPS™ behind the scenes to
help locate material discrepancies before offers are made.
GPS™ gives a Power-Assist for Quality Assurance
GPS uses an innovative highly intuitive data input interface. The user input begins
with eight simple questions used to tailor the
specific input screens the user sees. Although GPS structure has dozens of input
screens, the user generally sees only four to eight screens, depending on the answers to
the eight simple questions.
As the user enters data, GPS shows the user on-screen information as to whether each data
item is required or optional, and provides an immediate validation of the values entered.
After all data has been entered, but before the calculation phase, GPS checks whether any
inputs are outside standard ranges. The GPS Data Accuracy Screen report then
allows users to identify fatally invalid data in their inputs – including
errors that would not otherwise be detected until calculations are made, and possibly not at
all!
To streamline the data entry process, GPS uses separate files for Global Data and Case Data.
The Global Data are typically standardized and do not vary by case. Thus, the GPS user
avoids having to deal with this information repetitively.
GPS case data screens include Standard and Advanced sections. The Advanced screens include
data that rarely changes from default values, but the user always has access to the data.
Example: specifications for the optimized premium calculation.
GPS Interactive Valuation Summary Report shows not only a summary of results but also includes
the status of system errata indicators. If any errata are indicated, the user is directed
to an Errata Report to trace down the specific error that has been detected. As we learn
about new errors that humans in a hurry might make, we attempt to add them to the automated
error checking list.
GPS uses an innovative highly intuitive data input interface, designed so that users of
varying expertise and experience may comfortably use the level of assistance that they require.
Users are guided to the correct inputs as much as possible. Value Life Corp has programmed its
experience into the GPS software. GPS cannot assure that you always input the right data, but
it guides you forcefully and even precludes many common errors. The user input begins with
eight simple questions that GPS uses to tailor the specific input screens the user sees.
The Eight Simple Questions GPS will ask first
What is the illustration type? In-Force Illustration, or Sales Illustration.
Which applies to the illustrated premium? Varies 3 years (or less), Varies more than three years, or Carrier Minimum Premium Illustration.
Which describes the number of years illustrated? Normal Maturity, Extended Death Benefit Maturity, or Truncated.
What is the insurance product type? UL, VUL, Term, Non Par WL, Par WL.
What valuations are desired? Snapshot Valuation, Forward Looking Valuation(s), or Finance Collateral Analysis.
What health information is available? None, or Life Expectancies.
Is there a Verification of Coverage more recent than the Illustration? Yes/No.
Adjust Calculation Methodology to emulate another system? Yes/No.
Ask any providers, investors, or others about their review process for financial analysis
of life settlements. Almost universally, these users will agree that review is a good
idea, and virtually all will describe some level of review that they follow. Value Life
Corp wants to help you establish Best Practices for your review procedures. If your
first procedures generally produce values that stand up to review without error noted in a very
large percentage of cases, that is evidence that your primary procedures are being performed
well, but a strong review process is still needed. Even if the review process only finds
material error in a small percentage of cases, the financial impact of material error can
occasionally be quite large.
Value Life Corp believes that there are good reasons for strengthening the review process so
that the entity has the strongest possible
Independent Review
Process in place. Routine independent review can be implemented on a routine
cost-effective basis using the Q&A Program and GPS.
Independent Review Best Practices
- A Review is not truly independent unless it is performed by a different person
than the primary pricing analyst, and using an analytic tool different than the
software used in the primary calculation. These characteristics are required for
the strongest level of independent review.
- Every input item should be independently verified and all outputs should be reviewed.
- Automated routines and technology should be used to look for inconsistencies in data inputs and calculations.
- Material differences should be reconciled to the causative source. Reconciliation will not
only help in the current case assessment, but may also help to spot systematic approaches and conditions
that should be carefully monitored with all future cases.
Any person or entity that performs calculations relied upon by their clients to make purchase offers, or to advise
sellers about the reasonableness of offers, is always subject to some risk of error. Yes, most of these firms are
reputable and do the best job that they can with the tools and staffing that they have. It is prudent to consider
the risks that are assumed as to pricing accuracy and
what can go wrong. The main
point is that errors can and will occur. The key question is this: "
Do you have an adequate process in
place to catch errors on a timely basis?"
What Can Go Wrong / Some Sources of Error
- Bias or systematic error in the specific software used.
- Insurance product features not directly handled by the software.
- Clerical error in entering data to the input forms (i.e., the data is coded to the correct place but it is the wrong value).
- Mis-interpretation of illustration data and the proper place to code it on the input forms (i.e., the right numbers coded to the wrong place).
- Failure to correct for inconsistencies between the format of data on an illustration and what is expected in the input forms.
- Required missing data is improperly determined. Example: no illustration "as of" date, so the analyst uses the date the illustration was transmitted or printed. This almost always produces a material error.
- "Silent" adjustments made by the software to your input data.
Whenever quality assurance for pricing calculations is discussed with life settlement managers,
a wide variety of objections to strengthening reviews may be voiced. However, even if software
and staff training are beyond reproach, there is still value to having a strong independent
review process. Just having a well-designed review process in place can mitigate liability risks.
Ask your attorneys!
Objections and Responses
Objection #1: Our firm uses software developed and maintained by a well-known
and "trusted" firm that has many industry users, and the software is the most widely used
in the universe.
Our Response: You will be in good company then, and will have a good chance
for the same biases or even mistakes that many of your competitors will make, if you are
all using the same tool. In a "bid/ask" market place, you have exactly the same
information as your competitors. All software, no matter who is behind it, has the
potential to display some hidden "bugs" if you just hit upon the right combination of
circumstances.
Objection #2: Our firm has found that the software that we use tends to have
biases that systematically understate value, and we are comfortable with that.
Our Response: If you consistently buy policies for less than true actuarial
value, it will improve your returns. However, you may miss some cases in a competitive
marketplace if other firms are not following those same biases. There is an old saying
in the reinsurance business that you win your mistakes.
Objection #3: Our analysts are very smart and well qualified, and they review
their own work (or analyst #2 reviews work performed by analyst #1).
Our Response: Any review at all is better than none, but does it represent a
Best Practice? Often we see internal review processes that have a high contagion
factor where reviewers are sometimes drawn into checking off an incorrect input.
The review would be stronger if separate software, with different input forms is used.
Analysts often prepare as many as 30-40 cases per day, and it is not too hard to imagine an
occasional lapse. That is why our GPS software emphasizes automated review of data
inputs and calculation results.
In discussing Best Practices for independent review, we often hear the question "
How should
we resolve differences between the primary calculations and the independent review results?"
For starters, we recommend that a materiality threshold be established. For example, you might
establish a materiality threshold of 0.25% of face amount, or $10,000 if lower. Any differences
that are below the materiality threshold would be deemed to pass review, and a very large percentage
of cases should be in this category. If case differences exceed the threshold, then reconciliation
should be performed that determines specific sources that produced the difference.
Some level of difference may result from differences in the basic methodologies underlying different
software tools. Different software may utilize different underlying methodology. Although
most probabilistic calculation methodologies utilize conceptually similar projection techniques, there
are always some differences from software to software that will directly impact calculation results.
Differences between primary calculations and independent calculations (using different software) would
reflect the "noise" implicit in software-specific techniques. GPS includes
a mechanism to quantify
the noise.
Quantifying the Noise
GPS includes a calculation "switch" to emulate the techniques that have been observed in other software.
By running a first run with the "switch" turned off, the full extent of differences, including all
methodology differences can be seen.
By running a second run with the "switch" turned on (to eliminate the "noise") we generally find that
a very high percentage of cases run independently using GPS will show differences to the primary
calculations that are not material. It's the others that you need to look at.
Using this two tiered calculation approach helps to isolate systematic or input errors from differences
that are the result of intentional differences in methodology.
In the implementation of a Best Practices Independent Review Program, there are
practical
concerns for a manager who has the responsibility to run a business in a cost effective manner.
Practical Concerns Implementing Best Practices for Independent Review
Concern #1: We have to enter the data twice, the second time in the GPS software format,
and this will consume a lot of analyst time and therefore cost a lot of money.
Response: You can manually enter the data independently. However, GPS includes
an automated file conversion utility that will create the GPS input case data files from
Excel-based output provided by your primary software. This is turn-key and highly efficient,
and may be modified for different data sources.
Concern #2: If we use the data conversion utility, does that not skip a key area of
independent review?
Response: Yes, but you should consider the automated file conversion utility as a
first step. After a preliminary independent calculation using the automated file
conversion, you may find some material differences attributed to provable factors.
The independent review is strengthened by then checking the GPS case data files to the
original source.