Sometimes hundreds or thousands of personal property objects are destroyed or damaged and become the subject of a litigation dispute. The appraisal of each and every item of property can create burdensome cost and take considerable time.

Novotny has formed a business partnership with Dave Maloney and Richard Holgate called Large Loss Appraisals (LLA). The partners have adapted a statistical method to value large personal property losses. Other professions have long used statistical methods with established measures of reliability, but such a statistical approach has not been a part of typical appraiser training for personal property appraisers or a method they employ until now. Similar statistical methods are recognized and established in MASS APPRAISALS, commonly employed by appraisers to determine ad valorem property taxes for real estate.

Probability theory and a statistical method is often used by NASA, by many scientists and by many professions and industries. It is not commonly known that a statistical method is often applicable to very large personal property appraisal assignments with discrepant value claims for the same objects. In such situations a relationship exists that can be identified, calculated and used with known confidence and error intervals. Often in personal property litigated value disputes one party's claim (typically the claimant's) represents high retail (often new replacement cost) and the other party's value claim, typically determined by an appraiser, reflects fair market or used value for the same objects.

LLA has identified statistical equations that can measure the strength of that relationship and has imbedded the equation(s) in a powerful database that can be used, while employing the proper appraisal methods, to predict value confidently.

The LLA statistical equation was carefully researched and identified by University of California statisticians so that it would be relevant to the market shift ratio that exists when different markets are investigated by plaintiffs and appraisers. The primary object characteristics in the discrepant claim are the same; only the market level investigated to establish value differs. The statistical equation(s) necessary to analyze this market shift ratio were imbedded in a powerful database by a leading database expert. This necessary statistical methods and appraisal procedures are now being utilized by LLA and are available to all personal property appraisers and all interested clients.

The statistical method, proposed by Novotny, is based on random sampling. The statistical equation, imbedded in the database, calculates the standard error of the market shift ratio of the random sample, the sample correlation between the two value claims and then informs us when the sample size is large enough to meet the confidence and error intervals established. To test the proposed statistical solution Novotny proceeded to conduct a hypothetical study based on a completed appraisal of 400 pieces of jewelry that was part of a large loss assignment. To do so he:

- Selected a random sample (twice) from that population of 400 objects.
- Bear in mind that prior to the test the 400 objects had been competently and ethically appraised for both retail replacement value and for fair market value by two independent and qualified appraisers.
- In each test he hypothesized (pretended) that the fair market value of the 400 objects was not determined.
- He then input the fair market values that had been developed by the competent appraiser, one at a time and randomly.
- In each of the tests the equation informed him when the sample was large enough to meet the confidence and error intervals he established.
- He then had the database calculate the fair market value of the remaining objects in the sample of 400.
- In both cases, and in other instances, the Stein equation indicated that a sample of approximately 30 was sufficient..
- In each test the appraisal results (the FMV projections) were consistent with the actual values determined by the appraiser (in the aggregate)
- The confidence and error intervals established (95% confidence that the market shift ratio is predictably within 10% of true value) were confirmed in both instances.
- Novotny can confidently predict that if he repeated this test 100 times, that the results would be consistent (within 10% of the value determined by the appraiser) 95 % of the time.
- He would assume also that in 5% of the trials the deviation would exceed 10% of the value actually determined by the appraiser.

In other words, let us assume that the 400 objects had a total ** replacement value** of $100,000 as determined by a qualified appraiser. Let us further assume that a different but equally qualified appraiser's opinion of the value, defined as

In fact, in each of the tests he actually conducted employing the statistical approach and a random sample, the database rendered an aggregate value prediction that was indeed within +/-10% of the actual aggregate value determined by actual appraisal. An interesting feature of the database calculation that results is that highly discrepant predictions are highlighted thereby allowing for further investigations for those overly discrepant object value predictions by the appraiser.

The statistical approach is able to predict the fair market value of an entire population of comparable property types based upon the valuation of a much smaller random sample taken from that population. The statistical equation used is implanted in a powerful database in which all of the objects are documented. The statistical equation will analyze the market shift ratio for each object in the sample by comparing the fair market value of each object in the sample against the plaintiffs high retail claim for the same object. It will continue to do so until the sample is large enough to establish the relationship is strong enough to support the prediction.

There is a relationship between these discrepant values. Market participants react to the quality and value characteristics of a subject object in a similar manner since the object is the same object in both markets. Only the market level investigated by the plaintiff and appraiser differs. Thus a relationship exists that can be established and analyzed statistically.

LLA can reliably predict the value of an entire chosen population of similar property types based on the research and analyses of a limited sample of like objects randomly selected from that population for appraisal. This statistical methodology results in substantial savings in time and costs to clients while providing credible valuation results with a high level of confidence (95%) that projected values are within 10% of their true market value.

The appraiser must be competent, must properly identify the appraisal problem, competently define the population subject to the statistical approach and then identify and complete the scope of work necessary to develop credible assignment results for the random sample. The research and analyses should be based upon an investigation of an appropriate market level relevant to intended use and the definition of value that applies.

A personal property appraiser normally seeks to establish fair market value based on an analyses of actual completed sales (price agreements between informed buyers and sellers) of comparable properties in the most active secondary market. The opposing value claim, often prepared by a plaintiff, is often substantially higher since it is based on an analyses of a competing market level (often new retail price tags). Frequently a judge or jury must determine which market is appropriate or relevant to the legally mandated value definition. In any case a market shift is present due to the discrepant claims. This market shift can be calculated statistically using a MASS APPRAISAL approach.

It should be noted that many personal property appraisers have frequently used the market shift factor informally and intuitively, but without a proper statistical methodology. Such appraisers simply take a sample of some size, determine the fair market value of that sample, and then take the total of that sample and divide it by the total of the replacement value claimed for those same objects. It must be understood that the relationship of the values has not been tested statistically nor is there any basis to claim that the results are within established confidence and error intervals. In fact such an untested and simple approach does not calculate the standard error of the market shift within the sample to determine that a relationship exists or the strength of that relationship.

A MASS APPRAISAL methodology must be applied using a statistical approach for there to be any meaningful and defensible measure that establishes a basis for credibility. the extent of mathematical certainty that the results are meaningful and relevant to intended use must be identifiable and repeatable. MASS APPRAISAL is a standard approach, recognized in USPAP. The minimum standards that apply to a MASS APPRAISAL are established by STANDARD 6 in USPAP.

Personal property appraisers can now contact large Loss Appraisals to find out how to take advantage of this powerful statistical solution. In any assignment with two value claims for each object the statistical solution may, and likely will, apply. Any appraiser can utilize this powerful statistical method through Large Loss Appraisals, and still do their own research on the chosen sample. Of course there is a charge to calculate the predicted value that meets the requested confidence and error intervals.

Large Loss Appraisals has an office near Baltimore and in Los Angeles and Santa Barbara, each with a highly experienced personal property appraiser partner willing to travel when property can and should be inspected.

Find out more, contact Large Loss Appraisals now for substantial cost savings and a fast turnaround of appraisal results.

Visit the New Novotny Website

Call Novotny in Los Angeles area: 626-292-2224