Mercury’s ‘Research Analytics’ Breaks New Ground

If you' ve been reading the articles in this newspaper for a while regarding our revised Mercury Network, you might think that it will wind up being nothing more than a new AMC-like venture, or something akin to FNC' s AppraisalPort hub. But ordering and delivering appraisals will be just the tip of the iceberg.

With Mercury' s size and volume, plus the breadth and depth of the appraisers in our XSites Network (by and large, the cream of the crop of appraisal talent in the country), and the marketing and public relations weight that we can throw behind it (just as we did when we originally rolled out Mercury and landed over 235,000 ordering accounts), we' re confident that we can morph the Mercury Network into a true destination for real estate-centric information and expertise.

Part of that expertise will be funneled into what the industry calls "analytics". Simply put, given the complexity of today' s market, there' s a huge need for local, national, and property-specific analyzed data tailored to making decisions on individual properties or entire portfolios, or sometimes even broader policy. "Analytics" isn' t raw data, so don' t think we' re talking about harvesting data from reports. Instead, analytics summarizes information from a broad variety of sources, so that the client (any kind of client — lenders, homeowners, journalists, Wall Street analysts, etc.) can quickly and easily see the support for drawing conclusions about the real estate issue at hand, on any scale.

We' ve organized our analytics into three logical categories — "Order Analytics", "Report Analytics", and "Research Analytics". We' ll cover all three here, but as the title suggests, we believe the Research Analytics will impact the industry, and you, the most.

Order Analytics

In the case of a lender placing an order, it may be that they need to know beforehand whether it' s even wise to do so given the parameters of the property and loan amounts in question. (That' s commonly AVM territory, though AVMs aren' t the only source of such information and may in fact be poorly suited to it.) Most appraisers experience the impact of that sort of pre-ordering need for analytics as the dreaded "comp check". While it' s annoying to a professional who gets paid to render value opinions to be asked to give a pre-appraisal value opinion for free, and can' t be done per USPAP in any sort of cost-effective manner, that doesn' t mean that the underlying reason for demanding for such a service is unfounded.  It just means that the person the lender asked to get it, the appraiser, is the wrong source. But, there are better non-AVM sources of that data which don' t cause any problems for appraisers, and which in fact provide the appraisal profession a number of solid benefits.

Look for example at the role that a Zillow market report could play in removing comp checks. If you haven' t seen Zillow before, go visit it soon (www.zillow.com) and type in an address.  Everyone checks their own address first of course, and notes the things that Zillow got "wrong". Judged by appraisal standards, there' s no way anyone should use Zillow to determine market value. But combined with the comps data that it shows — in very visual, appealing satellite map format — there' s no doubt that a lender could see the activity nearby and similar to the subject, which is all they' d get with a comp check. The difference of course is that you didn' t have to do it. We can indeed provide Zillow-like and even Zillow-created information in a Mercury Network Order Analytics page, and we will.

Going beyond what you' d provide in a comp check, Zillow does a good job of showing zip code, area, MSA, regional, and national statistics, with (clearly now) a lot of declining trends. The data drill-down and zoom-out capability in a Zillow report is superb.

Another popular real estate site, Trulia, provides some overlapping capabilities as well as some unique views into property currently on the market which would be of value beyond a comp check. Trulia aggregates data by crawling agent websites, as well as by offering integration to agent technology providers so that the listings in an agent website, corporate franchisor, or even an entire MLS can be pushed (in agent market parlance, called "syndicated") into the Trulia database.  Our own Agent XSites syndicate their listings to Trulia every day, for example, with real-time updates as the listings change. Since others do the same, in areas where Trulia has good coverage, the listings data is a very fresh representation of homes currently on the market. Trulia also offers excellent "heat maps" of price and activity in areas, as does Zillow.

We believe, as an order is placed (or beforehand), a lender or other type of appraisal client may want to see similar area statistics going beyond just values or comps per se, giving an indication of an overheated or declining market, which could and should properly influence the client' s opinion of the loan application and the value indications from an appraisal, BPO, or AVM. (While we never suggest BPOs or AVMs, we' d love to see appraisers be in a position collectively to influence the confidence level that a lender might have in the results from them.) Since even the best appraisal is a value at a point in time, the surrounding market influences should factor into the lender' s view of the soundness of the loan in the future.(pq)

Currently, many appraisers are having reports called into question because values fell after the report was produced and the loan was closed. Lawyers dig into the comments and the dreaded checkboxes indicating market direction to determine whether the appraiser gave an indication that the value could change in the future. Of course, virtually every lender, borrower, agent, and loan broker assumed all along that it would change in the future – going straight up. But that' s conveniently forgotten as soon as a lawyer gets involved.

Had there been information pushed proactively to the client as the order was placed, either on screen during ordering or in a follow-up e-mail report where a web interface wasn' t used, indicating market trends and additional sources of data, as well as highlighting potential risk factors in that market, and laden with disclaimers and potential warnings, life would be a lot more pleasant for those appraisers currently under the gun. 

Now, make no mistake, no amount of data or disclaimer will ever provide legal cover for a rubber-stamp appraisal with boilerplate that doesn' t address "known or knowable" market realities.  But there' s no question that a proper appraisal now under scrutiny due to changed market conditions would be in better shape if the appraiser could point to a report preceding the delivery which advised the lender of relevant statistics and trends, and warned about potential risks which may be known or unknown. Having the report and the delivery receipt of it as part of the appraiser' s Digital Workfile certainly wouldn' t hurt.

Even outside the realm of legalistic CYA, the simple fact is that good lenders want and need that information, and even given the plethora of data out there, very few get it when they need it, and it rarely winds up in the hands of the person who needs to be seeing it.

That' s what Mercury Network' s Order Analytics will do. We' ll present and cull data from a variety of sources in graphical and tabular format — from other companies who specialize in market analysis such as Zillow and Trulia, as well as publicly available data such as census and assessor data. 

Likewise, we' ll include relevant traffic statistics summarized from our own server activity, such as map geocodes, based on our Mercury Appraisal Transaction Index, or "ATI".

The ATI is simple but effective. Each time an appraiser geocodes a street map in a new report in WinTOTAL, our servers know the latitude and longitude of the subject and therefore the zip code, census tract, and county in which the appraisal is being done. This happens many times per minute all over the country, many millions of times a year, since our geocoding servers handle over half of all the appraisal reports in the United States. 

In the larger counties and zip codes of the country, the ATI' s cumulative anonymous statistics over time show a solid trend of hot spot and cold spot activity, like the heat map from Zillow and Trulia, except as related to appraisal volume. (No values of course or other data are used, since we don' t have that data anyway and don' t want it. We just know the lat/long.)  

The ATI can show if transaction counts have gone up or down over time near the subject, which is especially useful in the short term – the 90 days or so prior to the report — since assessor sales volume data is so slow to be added to national databases.  

Mercury' s ATI will be the only one of its kind, and will likely remain so, since nobody else serves enough appraisers to produce statistically valid results. Will it be perfect? Of course not. But it' s better than the finger in the wind that dominates the decisioning process today, and it points a valuable spotlight on appraisers when it comes to real estate statistics.  (See the Research Analytics section of this article for more on the ATI and its use on a national scale.)

Report Analytics

Report Analytics, as the name suggests, are those included in the appraisal report delivered to the client. Whether it' s support for individual adjustments and metrics, overall market trends, neighborhood activity, or other graphical and tabular data, there' s a variety of valuable information you can put in an appraisal report — and directly affect the way the report is considered by the client.

Today, there' s an understandably large amount of interest in report analytics to support — and perhaps most importantly, present to the client in a pre-emptive manner to stem criticism — the derivation of negative time adjustments for the specific subject property. 

In this newspaper, in the WinTOTAL Yahoo User Group, and in the seminars we teach both live and via the web, you' ll see that two of our users in particular, Anthony Young and Patrick Egger, have been especially vocal about the need to include Report Analytics.  Both of them have been teaching the use of Microsoft Excel graphing and trend analysis for inclusion in appraisal reports, and both are featured in this issue of AppraisalPress. The latest installment of Patrick' s must-read series is on page 21 of this edition, "Down Cycling", and Anthony' s superb "Trend and Analysis Graphing" is on page 8. Also, we highly recommend Anthony' s Excel webinar series, at www.alamode.com/training.

Both are especially active on the WinTOTAL User Forum on Yahoo, where there are other appraisers as well with considerable expertise to share in the use of Report Analytics. We highly recommend that you join the Yahoo group if you haven' t already and learn more about how you might incorporate Report Analytics into your appraisals. In fact, there' s so much valuable general appraisal discussion there that it' s probably the most useful appraisal forum available anywhere even for non-WinTOTAL users. It' s not an a la mode-controlled forum, so you get uncensored information straight from fellow appraisers. At press time, there are 3,430 members, making it by far the largest online community of appraisers in the country.

You can join the Yahoo forum at http://tech.groups.yahoo.com/group/Total2000UsersGroup/ and learn much more about Report Analytics, which we don' t technically handle through Mercury itself.

We are, however, adding some Report Analytics into our WinTOTAL program. Members of our a la mode Labs group, including David Braun and Jim Simmons, have developed an automated spreadsheet system which produces superb analytics using MLS data (or any other data source). There' s an article in this edition of the AppraisalPress, on page 16, titled "Get The Market Conditions Analysis Kit In The Labs", where you can learn more about what the kit does and how it produces analytics, and download it for use in your reports.

Regardless of exactly how you approach Report Analytics, whether through your own spreadsheets and data analysis or by using tools you buy off the shelf, the simple reality is that you have to approach it in some way no matter what. There' s too much at stake today in an appraisal to not be able to show that you derived adjustments and supporting conclusions using sound methodology.

Research Analytics

Research Analytics are divorced from a particular appraisal or a particular order. They may be totally unrelated to a transaction, but they drive traffic and especially client awareness, which results in more transactions. Getting more appraisal transactions is good for all of us. 

In essence, any time you' ve seen cumulative statistics and trends regarding national, local, or regional real estate, you' ve seen Research Analytics. Think of how many times you' ve seen news reports citing the National Association of REALTORS' analysis of home sales volume and median price, or the Mortgage Banker' s Association report on loan applications, or the National Association of Home Builders' reports on home starts and building activity (and the more subjective but very useful surveys on builder confidence). Nearly every time you see them, it' s because those organizations put out press releases citing the analysis they' ve done. Other times, it' s because a Wall Street analyst, economist, business, or think tank did their own report using the summary data available directly from those organizations.

Now, think about it one step further: The organizations for REALTORS provide real estate sales analytics, mortgage bankers provide mortgage banking trends, and builders provide home building trends. Who provides home valuation trends? Shouldn' t it be an appraisal organization? It should, but it isn' t. Appraisers aren' t even considered the definitive national source for data on home values, which, frankly, is absurd.

There are home value statistics cited nationally and locally of course, all the time. NAR' s median price is a good example. Zillow is also often included in newspaper reports of value trends. And Case Schiller Weiss' s (CSW) national index is also referred to quite often — in fact, it' s become the de facto standard for newspaper and analyst citations.  

When a REALTOR organization, a property information website, and an AVM/index hybrid with data in only a few metro areas are the definitive source of discussion of valuation trends in the country, something is wrong. It' s unlikely that the Appraisal Institute will offer a solution. We intend to fix it with our Research Analytics, bringing appraisers themselves back to the forefront in discussions about national, regional, and local valuation trends.

Local Research Analytics

On a local basis, producing the definitive analysis of market conditions can have career-changing positive consequences. Markets are starved for this sort of information even more now than ever before, and good analysis of what' s near and dear to everyone' s wallet — their home — will land you right in the middle of all the free publicity you could ever get.

For a great example of local analysis done superbly and with unbelievable PR and advertising benefits, look at what Jonathan Miller has done in Manhattan (coincidentally, Jonathan is also a WinTOTAL user).  Here' s what New York Magazine said about him:

"He is the Encyclopedia Britannica, the Wikipedia of Manhattan real estate," says broker Suzanne Sealy. In the past year, 300,000 copies of his market surveys were downloaded. "He' s the only appraiser that has groupies!" jokes his business partner, John Cicero. Last month, a seminar with him sold out Avery Fisher Hall. (It' s the room' s only full house this year.)

As the quote suggests, it' s not only Jonathan' s skills as an appraiser and his market knowledge that garners him near-celebrity status — it' s the fact that he publishes quarterly analytics which define him as the local expert.  To see more about Jonathan and the analytics he produces, go visit his website at http://www.millersamuel.com.

Granted, if you produce market analyses in Manhattan, Kansas, you may not wind up on Fox, CNBC, or CNN like Jonathan. But you will wind up in the local news, and your name and analyses will be discussed at the local watercoolers.

National Research Analytics

The catch is, where do you start? And how does a Research Analytics report done in Manhattan, Kansas wind up affecting not only an appraiser' s local position, but the industry as a whole? How does that report have value added to it to affect the national value discussion?

We' re making that possible by providing a set of data collection tools and pre-built analytics which Mercury members (and therefore, XSite owners) will be able to populate with the most recent MLS data from the areas they serve.  Every MLS supports data downloads of some form or another, and that data is put into the spreadsheet-style analytical tools we provide.  Basic charts, graphs, and summary statistics are automatically generated, including "outlying" statistics which would be the natural subject of local discussion — such as rapidly changing volume or price trends.

You then simply add in the narrative information which any local would expect to see on those statistics and the market as a whole. Highly local information, such as reporting specifically on particular neighborhoods (maybe an upscale waterfront area, or the condo market), differentiates your analysis and adds to the newsworthiness of it, and Mercury' s Research Analytics module will support that.

The report you write is then automatically fed to your XSite and up to the Mercury national site. Press releases to the local media are generated and sent automatically as soon as it' s received and confirmed.

But that' s not enough, of course. As the title of this article suggests, the Research Analytics in Mercury take advantage of "Web 2.0" concepts of wisdom of crowds. That analysis done by appraisers from counties and metros around the country is combined into higher level analyses, aggregating the underlying summary data in each appraiser' s market report.

Using our national media contacts, we' ll be publishing the national statistics from these collective reports each month, directing readers to the Mercury website initially. They can drill down or zoom up to any level of detail needed, but at the most granular level — your analysis — they' ll be handed off to your own XSite. The traffic and recognition flows down to each of you, but with a national spotlight which once again puts appraisers in the forefront of the property valuation discussion.

After all, that' s the whole point. Just like other sites, we' ll have high level analytics data collected from assessor records and from third party services, but ultimately the national media needs to hear and think the word "appraiser" any time a discussion of home values is taking place. It can' t continue to be a news topic owned by other players in the real estate industry. Mercury' s Research Analytics service, combining appraiser-produced localized yet national reports with Mercury' s unique Appraisal Transaction Index, will finally make that a reality.