21st Century Economics

By comparing the CPI index level at two different times, you can make a statement about how prices have changed between he events. For instance, In December 1988 the CPI was 120.7 but December of 1998 it had gone up to 163.9. Doing the division of 163.9/120.7 =1.358, now subtract 1 and multiply by 100 to get a 35.8% rise in the CPI in the 10 years from 1988 to 1998 or about 3.6% per year rise.

The contents of the ” fixed basket of goods and services” is determined by the Bureau of Labor Statistics (BLS) after conducting a survey of consumer expenditures about every 10 years. The items being purchased rarely change but the BLS can adjust the weights of each of the 364 items in the basket. Some major changes were made in 1998 to the CPI.

The number of categories that the ” fixed basket of goods and services” is divided into went form 7 to 200 and the item structure and weights were changed. A more important change is that the CPI will not be calculated using a geometric mean estimator for about 60% of the expenditure categories that comprise the hypothetical market basket. The effect is subtle but important. It means that the quantities of goods in a particular category can change in response to the relative price changes. The new method of CPI calculations lowers the CPI value by about .2% over the old method.

In the past, the quantity was fixed and as prices of the items increased the CPI rose. The problem was that this did not reflect consumer spending. For example, if the cost of one vegetable rises, consumers will migrate away from that by buying a different one that has not risen as much. If the CPI reflected only the one that rose in cost, it would distort the picture of consumer inflation.

CPI’s value is that it is taken to be regarded as THE measure of inflation but because it is subject to consumer responses and handles the introduction of technology poorly, it sometimes results in a number that is larger than actual inflation by .5 to 1%. It also does not reflect stocks or government taxes which can have a major impact on the economy. Still, it remains one of the best indicators of inflation.

As noted above, the value of the CPI has risen by an average of 3.6% for the past 10 years but this does not reflect the month-to-month volatility of the usual method of reporting CPI which is to report the percentage change from the previous month or to report the CPI on a monthly basis.

For instance, the monthly percentage change CPI in April 99 was .7% but in March 99, it was .2%. You might interpret this to be a 350% rise in CPI from March to April when, in fact, the actual rise was from 165.0 in March 99 to 166.199 in April 99 - a rise of about 1.2 index points our of 165. A very small amount. Even if this were seen as the inflation rate for that month and you projected that for 12 full months, it would equate to an annual inflation rate of 8.7%. This gives a totally incorrect view of the economy and would be drastically out of line with the trend for the past 10 years of 3.6%. In fact, the entire year to date (January 99 thru May of 99) has a total rise in the CPI of .97% (less than 1% overall or an average of about .19% per month giving an annual rate of just 2.3%) or about 25% of the 8.7% figure). This is how the Wall Street wingnuts and economic pundits manipulate figures to scare people into buying or selling their advice and products.

What is important is how the market reacts to the CPI. If the CPI changes, in general the market goes in the opposite direction. Bonds (fixed income) and equities go down when the CPI goes up and vice versa. It is seen as less volatile than the PPI and so it is used as a better indicator of long term inflation trends.

Like most economic indicators, CPI does not provide proof positive of any particular trend but in combination, it can provide some insights into where the “trends” and “pressure” is pushing the economy.

Right now, the total picture, including the CPI, indicates that the although inflation is being constrained by heroic efforts by the Fed, we are seeing a weakening US Dollar, rising unemployment and very high energy prices that are adding to building pressures to push inflation higher. Employment compensation is rising rapidly, housing and vehicle markets are taking a dive, raw materials have risen and the Fed has lowered the rates several times with a high likelihood of a second lowering later in the year.

21s Century Economics takes the position proposed by Don Paarlberg in his book, “An Analysis and History of Inflation” (Praeger Publishers). In it, he studied 15 different economies from Ancient Rome to modern day Brazil and concluded that a moderate degree of inflation is usually accompanied by increased economic activity.

This has certainly been borne out by recent US history. He also proposes that inflation is not caused by production or prices but by the supply of money controlled by governments. A careful examination of what the US Treasury, M1 money supply and other currency exchanges were doing in each of the inflationary periods over the last 50 years proves that Paarlgerg is right.

The M1 supply adjusted for CPI plus stocks and government taxes (collectively referred to as the MCCPIG) has flattened or declined prior to every single one of the 11 economic downturns since WWII - with no misses or false alarms. The M1 supply rose 3.7% between May 97 and May 99 - remarkably close to the average CPI of that period.

A chart of M1 (deflated by the MCCPIG) since 1994 shows a percipitous drop from an index value of just under 170 in 1994 to under 100 in 2000. Given the Paarlgerg theory, the adjusted M1 (deflated by the MCCPIG) indicated that we were headed for a monster recession of epic proportions. At that time, Alan Greenspan said, “..storm clouds are massing over the western Pacific and heading our way”. Buy an umbrella! He and the MCCPIG were right on. In March, the bottom dropped out of the tech market and we entered a protracted bear market.

Personal income represents the compensation that individuals receive from all sources - wages, dividends, interest payments, proprietor’s income, transfer income (social security, welfare, unemployment) and other labor income. If we see this rise, expenditures often rise soon after. If Personal Income rises but expenditures don’t, then more people are putting money into savings. The nominal Personal Income and the Real Income (adjusted for inflation) are considered very good indicators of the current strength of the economy.

Increases in the PCE causes = The Stock Market to Rise
The Bond Market to Decrease
The Value of the Dollar to Rise

Decreases in the PCE causes = The Stock Market to Decrease
The Bond Market to Rise
The Value of the Dollar to Decrease

Bear Market Implications
People will spend money for three reasons:
(1) Because they have earned more
(2) Because they are buying something important
(3) Because they think that the value of money will soon decrease drastically

In both (1), you would see a rise in Income precede a rise in Expenditures. In (2) you would see a rise in Expenditures without a rise in Income. Often this would be matched by an increase in Durable Goods Orders. (See Guide on the Economic Indicator: Durable Goods Orders).
The same rise in Expenditures would happen in (3) with no rise in Income and often with a lesser rise in Durable Goods Orders. This is because when people are trying to expend money that they think will soon lose much of their buying value, they tend to buy consumables - food, gasoline, heating fuel, clothes, ammunition, etc.

If there are more people out there that think an impending Bear Market is real and will result in a major market setback, then you will see a large increase in Expenditures in the prior time period with no corresponding rise in either Income or Durable Goods Orders.

If on the other hand, there are more people that think that the economy will survive intact with little or no effects from Iran, Iraq, the new president or any other economic downtown, then the Expenditures will rise normal for the near term, and no rise at all for a seasonally adjusted Expenditure indicator.

If you see the Expenditures indicator abnormally rising, you can bet that people are stocking up on all the Bear Market kinds of supplies that all the doomsayers are saying will be needed after a major stock market setback - like moving funds out of equities and into gold or other cash equivalents. If that is the case, then you should be invested in stocks, bonds or commodities that reflect that potential.

If you see this rise, you can also expect that the value of the dollar will rise. If you see that, then buy gold as soon as you see a pattern of rise.

This rise in Expenditures may precede the actual rise in the value of the dollar and of monetary equivalents. The Expenditures rise should peak in just prior to an expected crisis (the indicator comes out between the 22nd and the 31st of the month) .

As soon as you see it, if it has risen by 10% or less over the previous money, then sell your gold. This is a cautious approach since you do not want to have to try to time your sell on a day-by-day basis by watching the paper or computer as the price of gold fluctuates.

This report introduces, in a simplified manner, the concepts of Activity Based Costing (ABC) as an introduction to the analysis applied to the Process Model and the development of the strategic plan for departmental analysis of the organization.. The department used in this example was the IT department.

Description of ABC

Activity Based Costing (ABC) is a technique that measures the cost and performance of activities and the products or services generated from those activities (Cost Objects). The resources, which are commonly reflected by the general ledger, financial statement, or object class codes are traced to activities based on primary and secondary methods of consumption. Activities are traced to cost objects, which are the functional outputs of the business processes based on their use.
The task of differentiating the organization’s activities as either value added or non-value added is perhaps the most important theme in ABC. Non-Value Added activities become candidates for elimination or reduction whereas Value Added activities become targets for improvement.

Traditional cost accounting systems do not provide adequate information to identify the causes of cost. In situations where costs are deemed by management to be too high, managers tend to rely on across-the-board overhead cuts to control spending in the absence of proper information. Thus, when funds decline or disappear, organizations usually respond by “tightening the belt” in the wrong way at the wrong point in the enterprise.

Common approaches include:

Universal reductions in the budgets of all departments;
Freeze on wage increases;
Freeze on overhead activities;
Early retirement;
Freeze on training and nonessential travel;
Freeze on hiring; and
Freeze on investments.

Such well-intentioned efforts generate a self-feeding cycle of competitive decay. They do not address the demand for overhead resources - the activities that keep people busy. There is a natural tendency for managers to cut expenditure on activities critical to the mission of the organization both in the present and in the future. Deterioration in the quality of service and pressures on an overburdened staff prompt renewed spending and overhead creeps up. The problem is that the fundamental causes of cost were not corrected.

The most common and least understood factor that touches off such a cycle is management operating with the wrong type of data - data geared to accounting rather than management.

How ABC was applied in this case

A baseline represents the inventory of business policies, practices, methods, measures, costs, and their relationships at a particular location at a particular point in time. The baseline also comprises a set of business processes that provides the context to an organization’s work. Activity Based Costing (ABC), often in conjunction with BPR modeling, pulls together all of these factors to enable decisions concerning the advisability, value, and difficulty of implementing various improvement alternatives.

ABC recognizes the causal relationships of cost drivers to activities. Cost drivers are the factors that cause work to be performed and in turn cause costs to be incurred (i.e. resources to be consumed).

The activity based management approach to cost management breaks down an organization into activities. An activity describes what an enterprise does - the way time is spent and the outputs of the process. The principal function of an activity is to convert resources (materials, labor, and technology) into outputs. Activity accounting identifies activities performed in an organization and determines their cost and performance (time and quality).

For purposes of developing an ABC model for the departments, a simple and effective activity based management system incorporating the following steps can be used:

1) Determine enterprise activities.

To identify the activities performed in the IT process a series of surveys and focus group sessions were held with each member of the IT department.

2) Determine activity cost and performance.

Performance is measured as the cost per output, time to perform the activity, and the quality of the output. As part of the interviewing conducted in Step 1, data was collected from each interviewee regarding:
number of transactions for each service area;
duration for performing each activity one time; and
information pertaining to the salary of each individual performing each activity.

Based on the interview results and detailed budget reports, all costs are able to be directly traced to specific activities, except for, miscellaneous expenses and computer supplies. These two categories were allocated to all activities.

3) Determine the output of the activity.

An activity measure (output) is the factor by which the cost of the process varies most directly. For each of the activities identified in the model an output was identified and quantified. These outputs provided the basis for tracing the activity costs to the cost objects.

4) Trace activity cost to cost objects.

Activity costs are traced to cost objects such as products and/or services generated by performing the activities. The best approach to take for identifying the appropriate cost objects is to view the services or products from the perspective of the end user or customer. This is the approach that was used for this cost analysis. The end result of this step is the determination of the costs of various time and attendance methods in the aggregate and on a per transaction basis.

5) Determine corporate short-range and long-term goals (critical success factors).

This requires an understanding of the current cost structure, which indicates how effectively operating activities deliver value to the customer. An assessment is then made based on these critical success factors as to which activities are non-value added and which are value added. Non-value added activities are those activities not providing value to the customer or to the business. These activities are candidates for elimination. Value added activities are by definition critical to the success of the enterprise’s mission.

6) Evaluate activity effectiveness and efficiency.

Knowing the critical success factors enables an organization to examine what it is now doing and the relationship of that action to achieving those goals. Everything a company does - or avoids doing - is measured against the short and long-term goals. This provides a useful formula on which to base a decision whether to continue performing or to restructure an activity. Also, cost control is improved by ascertaining if there are superior methods of performing an activity, identifying wasteful activities, and determining the cause of the cost.  

The Short and Simple Explanation:

You pay a fee for the “option” to buy stock at a future price so you can sell it now. In other words, you have the option to sell some stock now at a high price and then, at some time in the future, you buy it when the stock price has dropped.

The real neat trick is that you can wait until that future time to see if the price did, in fact, rise before you elect to sell the stock at today’s price. If the stock does not rise, you simple elect not to complete the transaction and all you forfeit is the cost of the option.

Typically, you buy the option and then wait to confirm that it is going to go down. If it begins, then you can elect to exercise your option and sell now at its current price of let’s say $110. Suppose you sell 100 shares for a total income of $11,000. Now you wait for the stock price to drop. When it reaches $85 per share, you buy 100 shares for $8,500. You sold the stock for $11,000 and bought it for $8,500. You made $2,500 minus the cost of the option.

Typically, the usual option buyer would buy much more than 100 shares. You can see that if you bought 10,000 shares, you would have made $250,000. The advantage is that you bought the 10,000 shares with money you made from the sale of the 10,000 shares. Sounds weird but it’s done everyday on Wall Street. You also have a low risk since if the stock goes up or does not change, you can elect to NOT exercise your option. You lose the fee you paid for the option but that’s all you are out of pocket.

Where this is most useful is when you KNOW that the stock will move down. In the case of a known political, economic or Middle East crisis, we often do KNOW that some stocks will go down and some will go up and then down. Therein lies your chance for profit.

You should read this next section but you can also skip down to the section marked Cautions.

The Longer Explanation:

As you might expect, it’s a little more complicated than what is listed above. Here’s how it really works.

You tell your broker you want to short 100 shares of ABC, Inc. at $110. This means that you are entering into an agreement with the broker to temporarily borrow 100 shares of this stock at $110/share for specified period of time. Technically, you are borrowing the 100 shares from your broker in order to sell them to someone else at the current price of $110.

The broker either has the shares in inventory or he borrowed them from a client or another brokerage firm. The sale is made and the shares are now in the hands of a third party that has paid $110 per share for them. At this point, you have not paid the broker any money but you do owe him for the 100 shares.

Now you wait. If the price of ABC, Inc. goes down, to $85, you then buy 100 shares of the stock at that price. You have now spent $8,500 for the 100 shares of stock. You now return to your broker the 100 shares of ABC, Inc. stock that you borrowed. You borrowed the stock at $110 and sold it at that price for $11,000. Then, later, you bought it back at $85. You made $25 per share in profit or $2,500. You sold the borrowed stock for $11,000 and bought it back for $8,500. Technically, you sold something before you owned it and bought it back after you sold it. Sounds crazy but that is what is called Selling Short.

Under some circumstances, it is possible to return the stock to the broker before you have to pay to buy it meaning that it is a paperwork drill until he sends you your $2,500 profit.

Cautions:

As with all stock transactions, there is a down side to this activity. Suppose the price of ABC, Inc. goes up to $125. You borrowed it from the broker and sold it at $110. Now he wants his stock back but the price has gone up. You now have to go into the market and buy 100 shares of ABC, Inc. at $125 per share or $12,500. You can then return the loaned stock to the broker. In this case, you lost $1,250.

There are ways to protect yourself from too much of a rise in price by using a ‘buy stop’ order GTC (Good Till Canceled). You decide that if the price of ABC, Inc. rises $5 you want to get out of the deal. You would place a buy stop order at $115. Then, if the price of ABC, Inc. rises to $115, you are assured that you will get out at about $115.

You may also want to get out of a short trade when you have hit a certain amount of profit. In this case, you would use a buy stop at you maximum loss level and a buy stop at your profit target level. This is called an either/or order. You are placing two orders to protect you if the stock rises and to take profit if the stock declines.

For the most part, brokerage firms do not place a time limit on the shares of stock they loan. This is because they make a commission both ways. And also, they want to keep the customer happy. There are some other rules and limits on this kind of sale but it has its rewards.

As you will see, selling short is a very useful technique when you know a stock or other investment will go down. What do you think will happen to all those defense contracting companies after this Iraq crisis all dies down? What do you think will happen to those defense contracts - like with Haliburton, if an anti-war president is put into power in 2009? What do you think will happen to GOLD after the panic passes about an oil crisis or a war with Iran? You can bet money that they will go down from their Bush Era highs. Make Money!  

This is contrary to the expectation of genetics which would seem to predict that the son’s inherited genetics would tend to be more similar to the father. Instead, what was happening was that each male person born in a family tended to contribute to the average for all males in society. That is, if one person in the family is tall then the next must be shorter to average out to the average height of the general population. Galton described this as “regression toward mediocrity” and went on to develop some very sophisticated math tools and techniques to do what he called “regression analysis”.

At first this sounds like it is a remarkable discovery but upon closer examination, it is just common sense. Let’s look at what would happen if this “regression toward mediocrity” did not happen. Let us assume that the expectation of the genetic inheritance was actually the predictor of the height of the son. This would mean that the occasional tall father would have tall sons. Unless they all grew to exactly the same height and then stopped, we can guess that the occasional son would be taller than his father. But if we follow the genetic expectation, that son would also have a tall son. If we extend this for a few dozen generations, we end of with lots of people dozens of feet tall.

This would also mean that there would not be a level average height for males in the general population but an increasingly taller trend that increases with each generation. Since that has not happened in all of history, there must be something wrong with the expectation of the genetic inheritance theory. Dr. Galton’s discovery does, in fact, apply to the general population but it has been found to apply to nearly everything that has an “average” value for some aspect of it’s description.

>>It should be noted that over the past few centuries, there has been a very slow rise in the average height in the general population - men and women - but it is due to an overall improvement in nutrition and health care rather than in genetics<<

Law of Math

Regression to the mean is a statistical phenomenon that is a fact of life in nature. It essentially occurs where the measures (for example the average heights of men) on the average regress toward the mean or average. The net effect of regression toward the mean is that the lower measurements tend to be higher, and the higher measures tend to be lower. It is important to note that regression is always toward the population mean of a group. That implies that there is a unique reference value, called the “mean”, that is an intrinsic part of every group of anything.

This intrinsic reference value, if known, allows you to define each and every individual in the group as being either above or below that value - above or below the mean or average of the group. The best example of this is in school testing of college students. Every student is tested and given a score that is relative to the overall average of the entire population that takes the tests. If you placed in the 10 “percentile” group, that means that you have a score that, on average, only 10 percent of the population gets. In this case, you are not being compared to getting a perfect score on the test but rather the comparison is against the highest scores made by anyone that took the test. This kind of test scores are called “grading on the curve”. The one student that makes the highest score is the “curve setter” and all the rest are scaled according to how many made each score so that in the end you have the familiar “bell curve” of scores for the entire population. That bell curve is the intrinsic reference value for that test and that population of students.

It has been said that regression toward the mean is a phenomenon that is similar to several everyday expressions such as “law of averages”, “things will even out” or “we are due for a good day after a string of bad ones”. And one that I would like to add is “it can’t possibly get worse (or better) than this!” Basically what all these phrases are saying is that “extreme experiences tend to be balanced by less extreme experiences”

Formal Math

Because regression can be applied to so many aspects of life and events, it is a highly developed aspect of mathematics called Regression Analysis. It uses some very sophisticated methods to look at what might otherwise be viewed as almost totally random data. I will not turn this into a mathematics textbook but I think it is important to understand that certain kinds of information can be very accurately defined with the precision of calculated numbers. Such calculations put relative quantities on the choices that we are faced with in our daily lives. When properly applied, they can be used to show us the favorable choice to make from among some very complex alternatives. For instance, in the lottery, betting, sports, politics and hundreds of other areas that require us to evaluate choices.

Below I have summarized some of the potential mathematic procedures that can be applied to the decision making processes. Don’t be confused or distressed by the complexity of these functions. You will see that, like the basic concept of regression, much of it is logical and common sense, once you know how to look at it.

Finding the Mean

If you have a lot of data and want to find the closest consistent pattern of the data, you can apply a technique called “curve fitting”. This is the basic function of regression analysis - to regress the data into a mean or average value and then be able to describe that average value in a formula. The result is a regression or prediction line or curve. It is called a prediction line because it can be used to predict the response to data you have not yet generated. For instance, the prediction line for a coin toss is 1 in 2 or 50-50. As we saw in the test runs of thousands of tosses, this prediction line could be a very accurate indicator of the response of future coin flips. In more complex regression, it is possible to determine the average response of medical studies, voter responses, accident and crime data or consumer buying patterns.

In the case of buying patterns of consumers, as the quantity of sample data (the number of products being studies) and the number of times each is recorded (number of buyers in the study), the accuracy of the prediction line improves. This gets so accurate that it become profitable for supermarkets to pay you, with discounts, so that they can get information about your buying patterns. They do this by getting you to register with their “buyer’s club” or with their “discount club” but what they actually did was get a lot of data about you and then record your every purchase so they are better equipped to market to your needs and appeal to your buying patterns.

Curve fitting or the creation of the prediction line is what the horse rasing bettor does in his head when he analyzes the past records for each horse before predicting which one will win the next race. Using regression analysis, that process can be quantified so that you have a number assigned to the chances for each horse in the next race. In each race, if you bet on the horse with the highest calculated chance of wining, you’d do better than the best racing bettors that ever lived.

Finding Multiple Independent Variables

How Good is your Prediction Line

In some cases, the real world data that you are trying to analyze appears to be very random and sporadic. What this means is that there is an average but each event or value might be very close or very far away from that average. In the stock market, this is called volatility and is the measure of how wildly the value of the stock swings from one day to the next. You can actually calculate this volatility using a technique called the Correlation Coefficient (CC). This is a value from 0 to 1 that says how close your prediction line is to the actual data. If the CC = 1 then you have an exact match and you can predict every single event with perfect accuracy. This might occur if you discover that each person in a particular store that buys diapers has a baby. An obvious conclusion but one that you can use if you are the store owner by offering everyone that buys diapers a coupon on bulk buying of baby food.

A value of CC=.5 would mean you have a 50-50 chance of prediction of the next even. This would be the case of a coin toss and it would not be that useful for betting. However, if you had a CC=.5 for data such as your chances of winning a large lottery prize, then you have a much more usable figure. The different is in the application of the prediction line to the alternatives of the response line - in other words, a 50% chance of a heads on a coin toss is not as useful as a 50% chance of winning the lottery. The Correlation Coefficient validates your ability to use the prediction capabilities of the regression analysis you have done.

Measure a Small Group - Apply the Results to a Large Group

There is a whole field of study called Statistical Inference that takes sample data and uses it to infer or predict what a larger group will do. This is the essence of the marketing analysis that is done with focus groups and public opinion polls. In fact, some very fancy math is used to determine the exact sample size in order to achieve a reasonable degree of accuracy. You can also decide on the degree of accuracy you wish to achieve (called the Confidence Interval) and then compute how many data points or samples you need to collect to achieve that degree of confidence in your resulting prediction line.

This aspect of regression and statistics is perhaps the one you have come into contact most often with and didn’t know it. Besides the focus groups and public opinion polls often used in politics, there are surveys and buyer pattern analysis that is taken on a small scale and then applied to a larger group. We sometimes call these “pilot studies” or “sample testing”. This is often the only method used in drug testing and yet it is used to “predict” the responses of everyone that will eventually take the drug.

If CC=0 then there is no more correlation between the plotted data and your prediction line than random chance would predict. There are, in fact relatively few such instances of analysis since it is now becoming more and more clear that even seemingly random events can be described by fancy formulas or sophisticated regression analysis.

In all cases, the regression must refer to some baseline or reference value. This value is held fixed or independent and then a second value is compared to it. The first value that is held fixed is called the independent or predictor variable and the second value is the dependent or response variable. In all our discussion, we have used predictor and response variables but have not called them that. For instance, in the coin toss, the 50% figure of heads or the 50% figure of tails is the predictor value. In our test flips we averaged 50.082% heads and 49.918% tails. These were the response variables. In the real world, the actual response variable may never exactly equal the predictor variable unless you spend a lifetime flipping coins. However, it should be noted that before we flipped a single coin, we could know that the RESPONSE of the flips would be VERY CLOSE to the PREDICTOR value - and it was.