The change in the local CPI during the first quarter of 2007 began with two deflationary months followed by the largest monthly jump since we began collection data in December 2001. The aggregate effect is an annualized inflation rate of 1.5%, which is substantially below the national annualized equivalent of 4.8% for the same time period. The local numbers have been driven by a variety of categories, but fuel prices and housing prices are the primary factors causing the fluctuations in the index. Fuel prices, including both gasoline and fuel oil, have grown dramatically during the quarter based on supply shocks including unrest in Nigeria and potential strikes in Belgium. However, refiners have taken advantage of these circumstances and OPEC’s drive to curtail production to reinforce the price increases has had a major impact as well. Finally, speculators in the spot and futures markets have perceived the advantages and they too have “fueled” the price increases. The local housing market rebounded in March from a dismal February in terms of building permits and housing sales, but the outlook is middling at best. There is still a substantial glut in unsold homes, although not the 8.1 months worth that exists at the national level. The outlook for housing seems to be a slow to very slow recovery as the unsold homes are sold or rented, but the impact of the housing sector is likely to be muted over the next several months. We do not perceive either a return to rapid price increases in this market, or further deflation. On another note, the most volatile component of our index seems to be women’s and girl’s apparel which swings wildly from month to month, despite that we collect only retail prices.
For the quarter, some very significant changes arose, but they were virtually the same as those that dominated 2006.4. Household insurance rates continued to rise, as did prices for meat, poultry, and fish and video and audio. Other than these, plus housing and energy, the other categories changed very little for the quarter.
Consumers and businesses are always more cognizant of price increases than decrease, so there is a tendency to expect higher rates of inflation than actually occur.
The outlook for the second quarter of 2007 suggests continued moderate price inflation, however, without much hope of a return to deflation. Food prices should remain relatively stable, but fuel costs will likely increase with the OPEC production cuts and the other reinforcing behavior from the other groups mentioned above. However, OPEC is very cognizant of the damage their cartel can do to world economies and the potential for a global recession, so they will only push so far. If the other groups in this very important market react to OPEC’s behavior in the same direction, fuel prices may not be that bad by September. Consequently we see the second quarter as one of moderation in price changes.
Overall, 2007.1 was a good quarter for Jacksonville. Retail sales continued to exceed national growth rates, unemployment was very low, and inflation was well below the rest of the nation. Local companies (particularly those with headquarters here) may not have seen as much upward movement in their stocks as would be optimal, but 2007 started out quite positively. If the housing market rebounds as many expect during at least the second quarter of 2007, and oil prices remain relatively stable, we see no evidence to change our perspective of strong economic growth for the entire year.
The housing market is an extremely dynamic part of the Jacksonville economy. This market is composed of two parts. The first component, the portion most people seem to be familiar with, is the purchase and sales marketplace. Here buyers and sellers come together in a dynamic market to buy and sell properties, unknowingly setting a corresponding price by a method known as arbitrage. This repeated selling and buying creates an overall market equilibrium, forcing houses that sit above this equilibrium price to sell much more slowly than their counterparts below this equilibrium price. A much less frequently researched aspect of the housing market is housing rentals. The rental portion of the housing market works in much the same way as the sales aspect of the market. Like its sales counterpart prices are set from the repeated arbitrage of property. Unlike the housing sale market where ownership of the dwelling is transferred along with the sale of the property, in a rental market the property in question is no longer the home itself, but the rights to that property for a specified amount of time subject to a given set of constraints. However, by-in-large, both aspects of this market exhibit extremely similar behavior in most circumstances.
One would like to think that since these two are part of the same general market their must be some sort of long run movement, or some interdependence between the two markets. If this is the case, and we state that the markets share similar qualities, can we infer that their movements are stochastic in nature or does a deterministic relationship between the markets exist? It is important to note that in posing the questions above we have not contemplated all possibilities. For all we know the two markets could share similar stochastic trends, and also contain a deterministic relationship as well. These possibilities just seem to be the most pressing questions in our search for answers related to the movements of these two portions of the Jacksonville housing market. If we assume that the markets trends are stochastic in nature, we are assuming that the movements of these two markets are generated from a purely random process, and therefore the directions of these two would be dissimilar. We can test this hypothesis with a co-integration test to determine whether or not these variables move together over the “long run.” To test the hypothesis of a deterministic relationship we can employ Granger causality testing to determine whether these variables cause significant changes in one another resulting in a deterministic relationship between the two markets. For those of you reading this article that have prior knowledge of statistics and regression analysis, right now you might be thinking that “correlation between two variables does not imply causation.” This is a very true statement and in many situations people try and draw these conclusions based on simple correlation analysis incorrectly. There exist a myriad of problems in trying to use a statistical technique to make conclusions about which the technique was never intended. Correlation-causality relationships between variables are perfect examples of this misunderstanding of fundamental statistical techniques. Granger causality does not rely on correlation techniques at all, rather Granger Causality uses the fact that time cannot run backwards. Consequently, if a lagged variable X can better predict the future values of Y, than that of the lagged values of Y, a causal relationship can be implied between X and Y. Thus, X Granger causes Y. This particular technique will allow us to determine if any interesting causal relationships exist between the housing sales market and the rental market. The main relationships we are interested in testing are between the number of housing units sold and rented, and also the average sales prices of homes and the average rental price, within the local Jacksonville housing market. With these four variables we can hypothesis six distinct pairs of relationships. However the main relationships we focus on are the ones between the average rental and sales prices and also the number of units bought and sold between the respective markets. When looking at table one, we can see clearly the different relationship possibilities between these four measures. If the variable pair displays a causal relationship with its paired entry, the pair is noted with an asterisk. The variable in bold is said to Granger cause the other variable inside the entry. There existed no significantly co-integration variables pairs within our sample.
From table one; we can see that there exist significant deterministic relationships between the local housing market and its rental counterpart. Upon closer analysis we begin to see that the causal relationships are not as one would expect. In general one would expect a direct causal relationship between the prices of the two markets and their volumes. What this means is that one would expect the causal relationships to be related to the prices of the corresponding markets, and also with the volumes of the markets, respectively. Cross cutting relationships across market segments is not something one would necessarily expect. Further interpretation of table 1 reveals that the housing sales market is a driving force behind the local rental market within the Jacksonville area. We can infer that housing prices Granger cause the number of units rented, and also the number of housing units sold Granger causes the average rental price within Jacksonville. This then causes reverberating effects within the rental market when taking into account that the number of unit rented also Granger causes the average rental price. This research has extremely far reaching implications for most aspects of the local Jacksonville economy. With housing becoming a behemoth of economic growth during the last six years, and the local Jacksonville housing market rebounding last month (with sales rising over 300, or 33% in March and the average selling price rising $24,000) it is safe to say that housing will likely continue to be a sound economic growth engine for the local economy even in the face of a slumping national market. The causal relationships that have been laid out within this research can help act as loose guidelines for housing investors who split assets between the housing sales and rental markets. An example will help to clarify the situation. If you know the local rental market reacts positively to housing prices then you will have an edge on the competition who does not know this fact, thus allowing you to reap the benefit of an improving housing sales market coincidently with a improving rental market.
The Jacksonville economy appears to be growing at a pace that would easily rival any other city in the state of Florida. However, the growth of a local economy is a very tricky thing to measure. One of the very rudimentary methods for easily estimating local economic growth is based on a jobs multiplier for each particular export industry. Most local economic growth is generated from the exporting industries that the economy possesses. Here the word export is meant to mean any good or service which the local economy manufactures for use outside of the local economy. However the estimates for economic growth derived from this process are often very misleading. The LEIP project has taken the stance that growth in our local economy is best measured based on how well the companies that make up the Jacksonville economy perform. During the tenure of the LEIP project, we have calculated two separate indices to take into account the subtle structural differences between the companies that make up the foundation of our local economy. One index takes into account only the companies who are based in the Jacksonville area (called “Local Headquarters”). The other index takes into account all the companies present within the local economy, excluding the prior mentioned companies (“Local Presence”). In constructing the data set the decision was made to construct the two indexes in the magnitude of the Dow Jones Industrial index. In essence the index generated is a Dow Jones equivalent; for the entire local Jacksonville economy. From this index we can test to see what economic factors influence the growth of the index as well as what factors present in the local economy have an adverse influence on the companies present within Jacksonville. After several variables were considered, the main relationship employed relates the stock of unemployed workers within the Jacksonville economy to how well the companies present in the local economy perform. It is safe to say that in most cases if the companies within an economy are doing well they will create a certain level of job growth. From this it can be deduced that there is most likely a negative relationship between the level of unemployment in an economy and the relative performance of companies present within that economy. Regression analysis applied to the data from January 2002 until the present suggests that a one percent increase in the level of unemployment within the local Jacksonville economy corresponds to a decrease in the overall local presence companies’ index of about 870 points (the local headquarters relationship did not seem to exist). If alternatively one increases unemployment by one standard deviation, the result is about a five percent drop in the overall local present companies’ index. Consequently, there appears to be a substantial negative relationship between how well companies in our local economy perform, and the level of unemployment within the Jacksonville economy. Furthermore is seems that about forty-five percent of that variation inherent in the local presence companies index can be explained solely by the unemployment rate in the Jacksonville economy, and by the value of the index in the previous period. From these results one can extrapolate that the performance of the companies comprising this index is a function of the unemployment rate in that given period t, the performance of the companies in the previous period t-1, and a term that nets all other effects that are not directly observable, denoted here as e. It is also relevant to ask whether the stock price index drives the level of unemployment, or is it the level of unemployment that drives the local presence companies’ index. Based on a statistical method known as Granger causality, (for those familiar with regression analysis, and general causality testing the results are presented at the end of the article), it clearly appears that it is how well the companies perform that “Granger” causes the level of unemployment in the Jacksonville economy, and not the reverse. An approach to economic growth, like the one taken by the members of the LEIP project, is an avenue that must be explored in detail, but with caution to temper optimism.
Pairwise Granger Causality TestsDate: 07/02/06 Time: 22:06Sample: 2002:01 2006:12Lags: 4
Copyright © 2016 University of North Florida1 UNF Drive | Jacksonville, FL 32224 | Phone: (904) 620-1000
RegulationsConsumer Information | Disability Accommodations