# A Bold Forecast For 2009 Home Sales

Posted by Joe Manausa on Monday, March 2nd, 2009 at 9:38am.

Today we published our 3rd Edition of the Tallahassee Real Estate Newsletter. If you are not already on our exclusive subscription list, you can join for free in the form on the right. All it takes is a name and an email, and you'll receive the most analytical report of the Tallahassee real estate market that can be found.

I will start today's blog by saying you will want to bookmark this page immediately. In January of next year, when the 2009 home sales numbers have been posted,  you will be able to come back here in order to seriously mock me or to wonder in amazement at the mathematical marvel that I will have become (through self-proclamation of course).

Today I will astound you by predicting the number of home sales in Leon County Florida for the entire year of 2009, and I will guarantee my accuracy to a 99% confidence level! How's that you say? There is a 99% chance that I will be right, and if my explanation does not bore you to sleep, I think you will agree (don't worry, there are pictures for those of you who hate math!).

### Using A Normal Distribution Curve And Standard Deviations To Predict Home Sales

I hit on an idea recently that I could utilize my current market research in order to predict home sales in the Tallahassee real estate market (Leon County, Florida) for the year 2009, and be statistically certain (99% certain) that the predictive range would be correct. In order to do this, I needed the following:

• The total number of homes sold every month from January 1991 through January 2009
• My 10 year old Statistical Analysis Book  I used during my MBA program
• Google, to re-teach me Statistical Analysis

To be completely honest, normal distrubution curves and standard deviation are much more difficult concepts for a real estate broker to utilize and explain than any other subject that I have covered in the Tallahassee Real Estate Blog. And while you're probably ready to jump to another page right now (please don't, it will be worth your while...), I intend on explaining this in as simple a manner as possible, and you will benefit from an early warning/prediction of the remainder of the year.

### Key Definitions For Real Estate Statistical Analysis

The standard deviation is kind of the "average of the average," and often can help us find the story behind the data. To understand this concept, it can help to learn about what statisticians call normal distribution of data.

A normal distribution of data means that most of the examples in a set of data are close to the "average," while relatively few examples tend to one extreme or the other. We can take a set of data (like home sales each month) and convert it to a normal distribution in order to analyze it.

Let's say we are looking at home sales in Leon County each month. We see that typically, we sell about 300 homes per month, but there are some months that go as high as 500, and some that go as low as 200. For the most part, home sales are normally distributed (meaning most month's sales are close to 300, while fewer months are much higher or much lower).

The standard deviation is a statistic that tells us how tightly all the data are clustered around the mean in a set of data. When the examples are pretty tightly bunched together and the bell-shaped curve is steep, the standard deviation is small. When the examples are spread apart and the bell curve is relatively flat, that tells you you have a relatively large standard deviation.

Computing the value of a standard deviation is complicated. But let me show you graphically what a standard deviation represents...

Notice the "bell shape" that we discussed earlier. It could be spread out more (meaning larger standard deviations) or taller and more narrow (meaning tighter standard deviations). The point to remember is that this is what we are seeking when assembling a normal distribution of data.

Even if you are a little fuzzy on what we covered thus far, hang in there! It will all make sense soon.

### Measuring Seasonality To Create A Data Set

By utilizing Tallahassee home sales data from each month back to January of 1991, we can create a measurement of seasonality for each month of every year. Such a measurement might look like the table on the right that has been constructed from Leon County home sales data from 1991:

### Creating A Normal Distribution Curve With Seasonality Figures

Ok, now we're getting to the good stuff. By creating a normal distribution curve of the January seasonality figures from Leon County home sales, were are able to determine ranges of mathematical probabilities of future home sales. For example, we know the following information:

• 144 homes sold in January of 2009
• On average, January sees 6.08% of the home sale each year

Therefore, by dividing the 144 home sales by its seasonality indix of 6.08%, we could say that January's home sales would predict (144 ÷ .0608) 2,368 home sales for all of 2009. Unfortunately, we need to be more accurate than that, and a normal distribution curve with standard deviations will allow us to be so.

Let's first understand how standard deviations give us different levels of confidence in our results. When we are only 1 standard deviation away from the mean (see the red area to the right), our level of confidence is lower than when we start getting to the tails of the bell curve. 3 standard deviations give us a 99% confidence (the red plus green plus blue area below) that all ranges of values will fall within there boundaries, while 1 standard deviation only gives us a 68% confidence that all ranges of values will fall within its boundaries.

Don't fall asleep now, we're coming to our conclusion...

### Tallahassee Real Estate Market Undergoes Statistical Analysis

When we create a normal distribution graph with January seasonality factors, the following real estate graph is created:

So, just to compare our results with the "standard normal distribution early, the area in red represents one standard deviation from the mean, the area encompassed by the red and green represent two standard deviations from the mean, and the areas encompassed by the blue, green and red represent three standard deviations from the mean. So what you say?

### Tallahassee Home Sales Forecast  with 99% Confidence

Our analysis yields the following

• With 99% confidence, we can say that home sales will be greater than 1,692 homes, but less than 3,942 homes.
• With 95% confidence, we can tighten our prediction to home sales falling within a range between 1,870 and 3,227 homes.
• With 68% confidence, we can tighten further to forecast home sales in Tallahassee to a number between 2,090 and 2,732.
• The 18 Year Average would forecast home sales of 2,368.
• My prediction: 2,675 homes will sell in the Tallahassee real estate market in 2009.This is a decline of 26.5% from 2008, which was the worst year on record....

So, where did my prediction come from?

1. My forecast for home sales in Tallahassee comes from a few observations:
2. Our bell curve is narrow, indicating a tight range of values for January seasonality (meaning January doesn't move around much).
3. There appears to be a recovery pattern whenever January measures in the 7+%, the following year is in the 5+% range.

So, with last year measuring a January seasonality of 7.36%, I am going with the established pattern and expect to see the same sort of recovery pattern. Using the ratio of the two previous patterns, I will use a seasonality factor of 5.38% (which yields total annual home sales in Tallahassee to be 2,675).

But time will tell.

OK, for those of you who made it this far, should I put this away and never do it again?

### Update 1/15/2010

Total Sales for 2009 were 3,250.

Joe Manausa, MBA is a 26 year veteran of real estate brokerage in Tallahassee, Florida and has owned and managed his own company since 1992. He is a daily blogger with content that focuses on real estate analytics and providing his clients with a tactical advantage in today's challenging market.

#### 3 Responses to "A Bold Forecast For 2009 Home Sales"

Big T wrote: Why not try Feb and maybe up to 6 months and compare results. Maybe Feb. or March are a better predictor of the year. Maybe a combined 1st quarter is a better predictor.

Posted on Monday, March 2nd, 2009 at 6:02pm.

Nick wrote: Joe,

Wow, as a homeowner, that's not a very pleasant picture at the top of the newsletter. It does make me wish I had some extra cash on hand to go house shopping, or at least to start looking seriously. This is a fantastic time for buyers.

The way I look at it is that it doesn't really matter for most people. If I sold my home now, I wouldn't recover everything I paid for it and put into it with upgrades, but the price on the home I would be buying would be reduced too, so the loss is really a wash.

Now, for first time homebuyers, this market is handing them the chance of a lifetime: they have the opportunity to enter the market at a depressed state - essentially a time flashback to the early 2000s (maybe, heaven forbid, the 1990s). Folks like me who bought two years ago, well that's when we hopped on this train; what I've lost from that I've already lost, but I'm really in no different position.
Fortunately, I love my house.

I'll read through this in detail later today, but as always, thanks so much for your regular detailed analysis and commentary.

Nick

Posted on Monday, March 2nd, 2009 at 6:04pm.

cliff wrote: Joe, although your statistical analysis is to be commended for the amount of time and effort put into it, the thought process is lacking vital information to put the real estate scheme of Tallahassee into perspective.
The saggging economy, unemployment, lenders rethinking their loan strategies and the new fanni mae and freddie mac regulations, not to mention how detailed the new stimulas program will be I would not want to be putting myself into the "forecast" role. Add the number of people who have lost up to 50% of their investments in the stock market and you have quite an unstable local real estate market. Even as the market as a whole comes back from it's over inflated pricing the 2-3% annual cost of living raise we have come to expect each year has now dwindled to maybe 1% wich translates into anyones guess on what is to come.

Posted on Sunday, March 8th, 2009 at 3:35pm.