Federal Funds Rate

I’ve been doing lots of thinking. I have lots ‘n lots to write about. For now I will leave you with this. The blue line is the Effective Federal Funds Rate (FFR) in percentages. FFR heavily influences the interest rate of everything. The red line is year over year percentage change of 35 to 44 year old employment level. Since employment level is arguably difficult to forecast, I added the green line. The green line is year over year change of 35-44 year old population in America, which is not hard to determine. I’ll write more tonight. I see the number of hits to the site and appreciate the interest, but I’d really enjoy some comments!


Texas Real Estate

I often hear how Texas was left out of the real estate bubble. I’ve heard a wide range of opinions as to why Texas’ real estate market was relatively robust of the past few years. I have a few opinions. First, Texas has high property taxes. I think this is due to Texas’ agricultural history. Either way, if you don’t have a high income you cant really afford an expensive house in Texas. The property taxes will overwhelm you. Second, Texas was the last state to legalize home equity loans in 1997. One of the provisions in the amendment was:

(6)  no additional debits or advances are made if the total principal amount outstanding exceeds an amount equal to 50 percent of the fair market value of the homestead as determined on the date the account is established;

Therefore Texas homeowners always had to have a large amount of equity in their homes to receive a second loan. Even after it was legal, Texans were slow to utilize home equity loans. Below is a graph of year/year change of Texas vs US home prices.

Spikes and dips in the 1980s had to do with oil prices.

The 1997 amendment.

2003 report discussing Texas homeowners uses of home equity loans.

2008 Liquidity Crisis


I found this paper in the comments at www.pragcap.com/qa-7.

The authors go into detail about how the liquidity crisis happened. They describe how banks work with the Federal Reserve, how banks took on the responsibility of creating acceptable collateral, and how they rejected their created collateral when they found out it was susceptible to “lemons.” It is a good read and further proves to me that my logic is incomplete but correct.

Age Distribution and CPI’s

I corrected the “US Change in M2, 35-44 Employment, CPI Housing” chart. I am now using Consumer Price Index for All Urban Consumers: Shelter minus Consumer Price Index for All Urban Consumers: All items less shelter. (CUSR0000SAH1 – CUSR0000SA0L2)


Collateralized Debt Obligations

A collateralized debt obligation (CDO) is a bunch of debt and/or other financial securities all grouped together to form another financial asset that can be divided into shares. The assumption is is that the risk of each individual asset in the CDO was accurately measured and accounted for with higher/lower interest rates. When you put it all together you get a distribution of returns with a mean and variance. The CDO is then divided into tranches based off a risk and return rate. The safest tranche gets paid first at the appropriate interest rate. The second safest tranche gets paid second at an interest rate slightly higher than the first, and so on until you get to the highest risk tranche who is paid a much higher interest rate but gets paid last.

CDOs exist to distribute risk. In the past you had to be a very large institution to buy enough debt to diversify, since you had to buy the whole thing. CDOs offered a diversified portfolio of debt that can be divided into shares. Due to their usefulness, CDO popularity grew rapidly. Globally, trillions of dollars worth of debt are placed in CDOs. They were considered short term assets and used mark-to-market accounting.

What went wrong? I am speculating now, since I do not work in the industry I have no way of knowing all the details. First off lets define what happened. In 2008, there was liquidity crisis and a run on the banks. Which caused the market to crash.

Okay, lets back up. In 2006, the Federal Reserve increased the Federal Funds rate. I think they increased the rates in due to the increase of GenY/Millennials/Baby Boomer’s kids employment level. As expected, some sub-prime mortgages defaulted. The returns of high risk tranches decreased, which is expected. However the returns of some lower risk tranches’ also decreased, which was unexpected. If you consider the assumptions used to calculate the risk and return of CDOs, it is clear that there is lots of room for errors, not to mention fraud. It is popular in financial media to blame fraud and greed as the cause of the crash. However I believe that there was always fraud and greed in the system, and that everyone didn’t become even more greedy or committed even more fraud in the 2000s. I admit that there was probably a tiny bit more fraud and corruption in 2000s due to relaxed regulation and increased in incentives, but the increase in fraud alone shouldn’t be able to bring a market to its knees.

Some tranches, that should not have been effected, lost some value, and caused everyone to slow down and do some due diligence. The due diligence turn up some bad apples, which caused the liquidity of all CDO to decrease further. Remember, most of these CDOs held to maturity would have paid out as expected, but a decrease in liquidity equals decrease in present value. Everyone’s balance sheet decreased because of mark-to-market accounting. For a bank on the boarder, the lost of value of their CDOs made them illiquid. Bear Sterns and Lehman Brothers became illiquid and bankrupt in early and late 2008, respectively. The decrease in liquidity of CDOs snowballed. Until there was a run for cash.

Was the bubble in housing really that big? I don’t think so. The 2008 market crash happened due to people panicking. An analogy would be the decrease in consumption of meat after a very small percentage of mad-cow disease was found.

Age Distribution and CPI’s

So, I have no idea if did this right. I wanted to normalize CPI of housing. I took quarterly year over year change of CPI for all Urban Consumers Housing and subtracted CPI for Urban Wage Earners and Clerical Workers: All items less shelter (CPIHOSNS-CWUR0000SA0L2). I think this shows housing price changes relative to other price increases. In other words, if overall CPI increased 10% in 1995, housing prices grew 10% inline with overall CPI. If over all CPI grew 10% in 2001, housing prices grew about 12.5%, 10% from my example overall CPI and 2.5% from the graph. The graph shows housing price growth above overall CPI. In addition I plotted year over year change of M2 money stock, which clearly shows a correlation with the change in 35-44 year old workers. It is intuitive to me that spending patterns depend on age, and that 35-44 year olds produce and consume more than other age groups. Below are two studying supporting this concept.

In the 1990s, this age group’s rate of growth declined. I think that without monetary and fiscal policy action, the demand for houses (and everything) would have significantly decreased, which most likely would have caused a multi-year deflationary contraction. The federal government has stated they will not allow deflation. Therefore monetary and fiscal policies were designed to cause asset price appreciation/stabilization. Their goal was to reduce price uncertainty, which is a requirement for economic growth. Here are some legislation passed in support of this goal (I am sure there are more):
Housing and Community Development Act of 1992
Federal Capital Gains Tax Rate History
Mortgage Rates (influenced by FED short-term rates)

I think their goal was to maintain stable prices and growth until GenY/Millennials/Baby Boomer’s kids (born 1980-1995ish) reached working age. Basically, real value (demand) started to decline in the mid-1990s due to age demographic reasons. The federal government wanted to keep nominal values inline with their selected inflation rate. Their plan was to slowly turn up interest rates as Millennials joined the work force. They started turning up interest rates in 2006. I think a mild recession was expected, but something went terribly wrong and caused the market crash in 2008. I have thoughts about what happened and will discuss it in my next post.

Either way, the point I am trying to convey is people matter. Economics is the study of people’s interactions. I prefer to skip the numbers and pay attention to people. During the “dot-com” boom, some famous investor said, “follow the silicon.” I say follow the people.

BTW, I don’t think asset price appreciation was the only reason M2 grew in the 1990s-2000s. M2 also grows when there are new valuables introduced into the economy, like… iPods and PCs. It also grows when productivity increases. Technological innovation is partially responsible for M2 growth, but I think it is relatively minor compared to asset price appreciation.

Discount Cash-flow Models

I start off by using a free cash flow to equity model assuming all equity and no debt, which is basically a free cash flow to firm model. I find the present value of operating cash-flows + non-operating assets.  Then I subtract out the current debt and off balance sheet liabilities and add back in the debt tax shield. I was taught to discount the debt tax shield with the risk free rate, but when I think about it, the debt tax shield is expected and sort of counted in as working capital. Therefore if firm looks like they run a tight working capital budget, I will discount it back with the same discount rate as equity. If the firm looks like they operate with extra working capital, I will discount the debt tax shield back with the risk free rate.

I use the current 10-year Treasury yield for the risk free rate, a 6 percent market risk premium, and the current 30 year Treasury yield for terminal growth rate. I typically use a three year beta (β) calculated with weekly data, but it depends on how much the business has changed over the years. I look at the 5 year, 3 year, 1 year, and 6 month β’s for any patterns. β is usually my only metric for market timing. It is a bit of intuition voodoo, and I could explain it, but it’d take too long, maybe later. I also unlever the β assuming a debt β of 1.0 unless I have reason to adjust it. I am inexperienced at evaluating distressed assets, so I typically stay away from them.

The main variables I forecast are revenue, earnings before interest, taxes, depreciation, & amortization (EBITDA), depreciation, capital expenditures, working capital, and debt. I carefully analyze historical SEC filings to determine the drivers of the changed, and then try to figure out the driver to the driver. I continue to find drivers to drivers until I am satisfied.

These are my definitions of the variables I listed above. The definitions are for companies in general. Certain industries may have different accounting rules or definitions. Revenue is a measure of the size of a company. If revenue grows it means the business grew and vice-versa. I believe there is only a few fundamental reasons for revenue to change:

1. Market size
2. Products line and/or upgrades
3. Price
4. Inflation

Changes in market size include things like competition, target demographic change, expansion in territory, ect. Once you have isolated why the revenues are changing, try to figure out what are the drivers of that change. There is almost always an additional cost to additional revenue. It is important to note if that cost is fixed or variable expense. For instance, Starbucks increased y/y revenue due to increase number store locations. The driver of revenue is store locations. The driver of store locations is a study on a specific demographics and coffee shop density. Additional store locations require capital investment in the prior year.

EBITDA measures profits after accounting for variable cost. It isolates cost to cost-of-goods, sales, general & administration SG&A, R&D, ect. If revenue changes you expect cost to change by the same percentage, thus EBITDA would remain the same. If EBITDA changes then the variable cost of the firm changed, and they are either becoming more or less efficient. Changes in EBITDA could be increased/decreased product prices while cost remain the same or the opposite.

Depreciation measures fixed cost of operations. Depreciation accounts for capital investments made in some prior reporting period. It usually calculated based on time (age) of the capital equipment. In some cases depreciation maybe estimated differently. For example, a truck’s value may depend more on miles driven than on time. Extending the truck example, EBITDA would be revenue minus the cost of gasoline and wages of the driver. Depreciation would account for the lower value of the truck.

I separated capital expenditures (capex) into two categories, maintenance & expansion. Maintenance capex is investments required to maintain current revenue. Expansion capex is investments made to capture additional revenue. In other words, maintenance capex is replacing your current truck after it falls apart. Expansion capex is buying an additional truck to increased revenue. A good estimate of maintenance capex is depreciation. One ratio to check if you are forecasting capex correctly is return on assets (ROA). You should question if ROA changes significantly.

Working capital is a measure of short term financial efficiency in my opinion. Assuming no interest or penalties, you want to get paid first and pay everyone else last. It also sort of measures the logistical efficiency of an operation.

I project debt, because most firms have an optimal debt level. Additional debt means more assets to generate revenue. Less debt means negative cash-flow.

Work Equals Returns

To start off I want to prove to you that generating risk-adjusted returns above the market  (α) is possible. I had a professor (in a non-financial subject), who truly believed the market was highly-efficient and beating the market was impossible. I could not disagree more. Indexes are average returns of a large and effectively static group of equities, therefore by the central limit theorem if you randomly choose a sufficient (about 30) number of stocks over a sufficient amount of time (arguable amount), it is statistically just as improbably to under-perform as it is to out-perform the average. Below is a graph proving my point.

You can see from the graph that α’s are more-or-less distributed normally. The graph is limited to data of funds with 10 years of continuous operations, therefore there are lots of funds that went out of business, but there are also lots of funds that did well and quit, or just didn’t have 10 years of records, or made a killing in management fees. Either way it makes logical sense that half the money out-performs the market and half the money under-performs the market. These funds are run by full-time accredited institutional investors with infinity more resources and operate infinity more efficiently (favorable brokerage fees) than an individual, but they also have constituents they must follow. Such as selling at the wrong time due to client demands, accountants, auditors, insurance, marketing material, ect. Beating the average is not going to be easy, but it is completely possible.

That being said I am not an accredited investor giving professional advice. I myself only self-invest a portion of my savings. Over time I hope that portion becomes greater due to my abilities to out-perform the market. My strategy has no rules really, besides thinking. I put as much effort into my sell decisions as I do into my buy decisions. I set price targets, identify weaknesses in my analysis, and/or reason to revoke my assumptions. I know that brokerages fees are a weakness and focus on minimizing turnover. I never make any assumption blindly. I read financial media for entertainment. I am realistic and know there is no way I can follow 30 companies at the moment. I know plenty of people who got rich in ways I don’t understand. I make lots of mistakes, and I keep track and analyze them.