Slow as Molasses Housing Market Affects Title Insurance Profits

According to a new report by AM Best, title insurers are experiencing losses as an industry for the first time in 17 years. This translates to a $84.5 million loss.

Factors contributing to the loss are similar to the issues raised in my last post on title insurance and include a large slowdown in sales of title insurance as properties languish on the market ( 10.4 months of inventory on the market, 33.1% drop in housing starts). Also contributing is inadequate loss reserving practices earlier this decade and agent fraud and embezzlement.

Though expenses are rising, the increase is due to falling premium revenues rather than a real increase in expenses. In my opinion, this is due to increasing levels of title recording automation at the county level, allowing title insurers to hold the line on their expenses due to running the title plant itself. Keeping expenses in alignment with premium revenues is a key challenge for title insurers.

The outlook for title insurers is iffy with the continuing housing adjustments.

You can see a free excerpt from the report or purchase the full copy.

New Credential For Kim

Just a quick self-congratulatory post to share with you. I have a new credential - the Certified Associate in Project Management. The credential is the lowest level that the Project Management Institute confers and involves having 1500 hours of experience being a part of a project management team (or 23 hours of education on project management) and taking a 3 hour exam on the material in the Project Management Body of Knowledge manual.

My project management experience is primarily the project I did at CNA in the 90's building the system and methods of measuring the asbestos, pollution, and mass tort exposures.

My purposes in sitting for this exam and getting the credential was to (1) learn more about project management for the benefit of my clients and (2) perhaps it will give me an edge when I bid on USDA crop insurance projects.

So, congratulations to me!

Project Management Concepts and Insurance Companies

Many elements of project management (PMI.org) could be applied to help insurance companies manage their operations and projects better, but one in particular struck me as being a strong area for most actuaries to help with. That area is the CONTROL area.

Control is the mate of Planning, another area that strong actuarial help is valuable. Control gets a bad rap from people that feel that the term implies heavy-handed oversight or punitive measures, but is, in reality, an essential part of the planning process. Where planning says what you are going to do and when, control checks on those plans and can suggest corrective action when plans are going awry, as they so often do.

There are three types of control processes - (1) Feedback Control, (2) Concurrent Control, and (3) Feed-Forward Control.

Feedback Control is frequently done by actuaries in insurance companies. Studies of rate levels, unallocated expense allocation, projected expenses, loss reserve calculations are often purely or nearly purely Feedback Control, in that they base the future projections on historical results. In project management, Feedback Control is considered the least optimal control method, since the undesirable events have already occurred well before the control function is initialized. Interestingly, projections based on historical events is quite favored by insurance departments of insurance, that frequently require explanation if your projections are based on anything else.

Concurrent Control is a type of control that takes place when a process is about to occur. The final checking before sending rate filings, Annual Statement reports, agent's bonuses, large claim settlement reports, etc. are are form of Concurrent Control. Training and periodic checking of employee work is also considered Concurrent Control.

The last form of control is an interesting one. It is a form of control more and more actuaries are getting involved in. This type of control function is called Feed-Forward Control and it's goal is to inspire corrective action before a deviation in the plan occurs. Actuaries call it "modeling" and have applied the concept to forecasts of future profitability, premium levels, claim costs for certain lines of business, enterprise risk management, or dynamic financial analysis.

The steps of Feed-Forward Control are as follows:

1. Identify all relevant input variables. For project management, these variables relate to time, volume, and money (or costs of the project). For actuarial work for insurance companies, the variables might be premiums, loss costs, expenses, investment income, risks of various company functions or operations, etc.
2. Build a dynamic model representing the process and keep it updated.
3. Collect data and enter into your model. For most projects I do, the data gets collected into the model before, during, and after it is built.
4. Perform regular assessment of the projected path and the variation from the plan. Are the loss costs out of range? Is persistency decreasing off planned values? Is the level of risk for a type of insured/coverage/policy unacceptably high? Are the number of claims or policies off plan?
5. Take action - Feed-Forward Control provides the early warning system needed to take action before getting too far away from plan.

Actuarial work for insurance companies generally uses the Feedback Control type the most, but Feed-Forward methods have the potential to change the profitability of the company quicker and more reliably. The major disadvantage for insurance companies is the more complicated filing and approval process for applications that require state approval, but that is overcome with clear actuarial memos and an education process. The advantages of getting ahead of deviations to your insurance company's plan outweighs difficulties getting the process approved. Rating agencies also appreciate the use and application of Feed-Forward Controls both for the added stability of the company and the evidence of careful planning that is implied in the use of such a method.