Money matters

Risk aversion

admin — Wed, 09 Dec 2009 21:49:24 +0000

Daniel Bernoulli, a famous mathematician. proposed a theory of utility in 1-38 that distinguished between price and value. where price is equal for everyone but value (utility) depends on the individual making the estimate and their circumstances.
Bernoulli’s approach defined a concept of diminishing marginal utility, which indicates that as wealth becomes greater, then the preference for more wealth diminishes. In the lower left part of the chart, where the investor has a Small net worth, the likelihood of accepting risk is much higher. although the magnitude of the risk is still small. When risk becomes greater, even proportional to rex., ard. all ins estors become cautious. Bernoulli’s graph shows the curve beginning at zero and mov ing up and to the right in a perfect 1/4 circle, ending horizontally, where risk is no longer attractive. This implies that people are risk-averse. Most people are not interested in an ev en chance of gaining or losing an equal amount. Other theories that have been proposed are that tile market maximizes the amount of money lost. and the market maximizes the number of losing participants. All of these concepts appear to be true and are very significant in developing an understanding of how the market functions.

Risk Control

admin — Fri, 04 Dec 2009 21:44:54 +0000

A trading system alone will not assure success without proper risk control beginning with each trade and continuing until a portfolio of different trading methods is created. Systems have losing streaks that will ruin any investor with inadequate resources and poor timing; a speculator must decide the initial capitalization. the markets to trade, and when to increase or decrease leverage. There are risks that can be controlled or reduced, called systematic risk, and another called market risk, that can take the torn] of a price shock and can never be eliminated.
This series of posts tries to cover a broad range of topics relating to risk, including individual trade risk, leverage, portfolio diversification and allocation, price shocks. and catastrophic risk. It is not possible to say that one is more important than another. in a specific situation, any one of the areas discussed may be the answer to preventing substantial loss. The first part of this series of posts discusses capitalization and shows why man,.traders m-ill be successful for months and then lose everything in only a few clays. It will explain the choices in leveraging and offer alternatives of less risk. The last section analyzes when a system is performing properly and when it is not living up to its expectations.
Risk control begins with a trading philosophy the most common is called conservation of capital. it is the assurance that the investor has been given the most opportunities for success, which usually translates into keeping losses small. This is often accomplished by allowing only small losses per trade or using a trend following system. Once a trend position is established, it is held as long as prices continue in the direction of the position it is closed out when the trend changes. The resulting performance profile is one of’ more frequent small losses and fewer large profits.

Variance, Option Value, and Technological Choice

admin — Tue, 01 Dec 2009 21:37:12 +0000

An important insight is that if you own a real option, variance helps you: If the price had been $28.09/MWh constant, you would have earned $10.8 million and you would never have shut down or reopened the plant. But because the price was highly variable around $28.09/MWh, you would have earned $28.6 million. It is your ability to “operate only when desired” that has value. Intuitively, “the bigger the upside,” the better for you. “The lower the downside” makes no difference: you are not operating anyway. Variability is on your side!
This does not contradict our intuition from the investments section. There, we posited that you disliked risk (at least systematic risk), and would only take it on if you receive extra expected rate of return. You may still intrinsically dislike risk, and require a higher hurdle (discount) rate for investments with much embedded real option values—such cash flows are typically very risky. But, in the presence of a real option, the risk also increases your expected cash flows—and often so tremendously that you end up much better off with risk than without risk. The present value, taking the higher discount rate into account, can be much greater.
Real options generally arise from your flexibility (here, whether to operate or not to operate). Different technologies have different real options and to different extents. For example, nuclear power plants have higher upfront fixed costs, but lower marginal costs. A nuclear power plant may cost over $1 billion to construct, but it may be capable of producing electricity at costs as low as $5/MWh. Therefore, nuclear energy plants typically run continuously, regardless of electricity price. They have fewer embedded real options. This has another consequence: If you ignore real options, you will mistakenly end up with too many high fixed-cost, low variable-cost technologies. You will believe nuclear power plants are much better than turbine gas plants, even if they are not.
Indeed, many economic resources are nothing but real options. Much R&D, e.g., into a new cancer drug, will never pay off in and of itself. But, if the drug development were to succeed, the pharmaceutical company would create factories and earn billions of dollars. An investment of R&D can thus be considered the purchase of a real option. Similarly, undeveloped land has zero inflows today, and still requires the payment of real estate taxes. Its only value is the real option to build on it if demand for land use were to increase in the future. And, your degree and next job may have more value, because you can walk away from them if other, better opportunities were to appear.

Making Money with Hindsight

admin — Fri, 27 Nov 2009 21:36:19 +0000

We should already be aware of the problems of back-testing. While there are no other alternatives for validating a proposed trading strategy, it is necessary to look at the problems more carefully to see a solution. No market is more evident than crude oil during the Gulf War, discussed in the previous series of posts. From August 1, 1990, through the end of the War in February 1991, oil prices were driven by news. But not all news is a surprise. In the agricultural markets, a crop freeze in orange juice or coffee is often anticipated by a change in weather. Before a freeze can occur, the temperature must drop. And that low temperature must he sustained to cause damage. This weather change causes processors to protect themselves by buying forward contracts or futures. In turn, prices move up.
Because Iraq had been moving troops near the Kuwait border, their intentions were not a surprise; however, diplomacy failed. The threat of an oil supply disruption caused oil prices to slowly rise. Many systematic traders and commercials would have been long on August 7, 1990, when Iraq invaded Kuwait.

PRICE SHOCKS

admin — Sun, 22 Nov 2009 21:34:54 +0000

Price shocks represent the most significant obstacle in the effort to close the gap between test results and actual trading, or expectations and reality. A price shock is, by circumstance, an unexpected event. Exceptionally volatile price moves are caused by actual news that differs from expectations, such as the Fed raising rates by 1% when only VA was anticipated; or, it may be a significant, unexpected political event, such as the abduction of Gorbachev in 1991. The key word to remember is “unexpected.”
If an event was expected, then there would be no price change. The market would have already moved to the anticipated level. Therefore, we cannot expect to profit from a price shock by clever planning, but only by chance. We should never assume that we would be on the right side of a large, unexpected move in more than 50% of those events. Unfortunately, when we back-test a trading system using historic data, we tend not to identify specific price shocks and treat them as normal, predictable events. We choose systems that perform best over a set of parameters, without regard to specific trades that may have been the result of shocks. We judge results by higher profits, lower risk, or a combination of statistical values. The results chosen as the best performance often have been the greatest beneficiary of these unpredictable price shocks.

Business risk

admin — Tue, 17 Nov 2009 21:34:01 +0000

Business risk is the uncertainty of income flows caused by the nature of a firm’s business. The less certain the income flows of the firm, the less certain the income flows to the investor. Therefore, the investor will demand a risk premium that is based on the uncertainty caused by the basic business of the firm. As an example, a retail food company would typically experience stable sales and earnings growth over time and would have low business risk compared to a firm in the auto industry, where sales and earnings fluctuate substantially over the business cycle, implying high business risk.

Systems That Work in Only One Market

admin — Fri, 13 Nov 2009 21:32:52 +0000

From time to time, all traders receive mail offers for a highly specialized system called Cattle Trader,” the “Silver Day-Trading System,” or “Easy Profits through Stock Index Trading.” It is most likely that each of these systems has been finely tuned with rules unique to this one market. Once aware of the optimization process and its results, these offers must be viewed more critically; after all, with optimization techniques able to test combinations of indicators and rules, you could just as easily create a system that seemed to make as much on a single market. To prove that you have actually found something that works requires an understanding of the rules and the way it was tested, or time to watch the System operate in the real market. If possible, a comparison of the real trading profile compared with the expected performance based on testing should be available.
Student t-Test
Among tests that help determine whether the results are significant, the student t-test is one of the most useful. It tells you whether there is a systematic bias in the data by showing whether the mean of the data is significantly different from zero. This is a useful piece of information when you try to decide whether the results of only a few trades represent a good system, or whether a series of losing trades implies that a system has no value.

Chaotic Patterns and Market Behavior

admin — Tue, 10 Nov 2009 21:30:47 +0000

Chaotic patterns are easy to imagine in the behavior of prices, but very difficult to measure. There would be no problem in predicting price direction if every participant reacted in the same way to the same event, much the way a single planet would smoothly orbit a single sun. In the real world nothing is quite as simple. Consider the pattern of prices represented as planets that are affected equally by two events, E I and E2. We will get a wobbly pattern whenever the planet passes across the midpoint where one attractor is stronger than the other. At point a, the object is most allected by the nearest attractor, E l, but as it circles it becomes closer to E2 and tries to form an orbit around it. The possible patterns are too complex and they vary based on the distance between E 1 and E2 and the size of E2 compared with E l. If attractor E2 is much larger than E l, there will simply be a distortion in the orbit around E2; if E 1 and E2 are the same size, objects a and b will switch orbits, forming figure eights.
As complex as these patterns might get, they are simple when compared with reality Each day brings events of various importance into the market, acting as attractors. Each attractor has an initial importance that loses value over time. To make matters worse. we cannot predict when a new attractor, or news event, will appear. This makes the chaotic pattern very similar to raindrops falling on a pond. Each new drop. equivalent to a new event, hits at an unpredictable time and place, with variable size, and forms circular ripples. These ripples dissipate as they get further away from the point of contact in the same way that the importance of an event fades away over time. The interesting aspect of the raindrop analogy is that, while we cannot predict where the next raindrop will fall, once it has landed we can completely determine its effects-until the next drop hits. This is remarkably similar to the market. Less often there is a significant event, a price shock, that overwhelms the smaller events, the market noise, for a short time.

TUNING TWO TIME FRAMES TO WORK TOGETHER

admin — Mon, 09 Nov 2009 21:30:32 +0000

Throughout most of this series of articles, the individual systems and methods have been discussed for their own merits. An analyst would look for the specific RSI or stochastic that somehow generated the most profits by showing a trend change or an overbought, oversold condition. That is not the optimal use of an indicator when it is expected to tell you the best time to enter a trend. instead, you will want the time period for the indicator to be much shorter than the time period used to calculate the trend. For example. if your trend system has a typical holding period of 1 month (about 23 business days), you may want your timing oscillator to have a good chance of giving you a better entry point within the first 5 days (20%) of that trade.
To get an indicator to reach overbought and oversold levels an average of every 5 days. you will need to use less than 5 days to calculate the indicator value. it doesn’t matter if the indicator returns a loss when the results of all its buys and sells are totaled, because the profitable part of the program comes from the longer-term trend component. It is most likely that the oscillator will generate five overbought or oversold conditions during the typical holding period for the trend trade, but only one of them will be used-the first one after the new entry signal. It is only important to see how the timing helps to improve overall profits and risk.

Multiple Time Frames

admin — Mon, 09 Nov 2009 21:29:17 +0000

Although the use of multiple time periods for analyzing markets has been popular for decades, few professionals have talked about it. It is only since better quote equipment has allowed this technique to he accessed by a wider audience that this approach has begun to appear in the public domain. The combination of multiple time periods allows the trader to time entries into the market using very short-term data. such as 10minute bars, while watching the longer-term picture for the daily or weekly trend. Because it is agreed that most trends are best identified over a longer time period, while choosing the entry point requires a much faster response, the combination of two, or even three time intervals is very sensible. if the trend can be identified profitably, then the trader can filter or select short-term trades that have a better-than-average chance of becoming winners.
For most traders, the use of any one time frame presents special problems. The very short term contains a high percentage of noise and obscures the market direction. The attempt to isolate patterns within charts of 5minute bars can divert Your efforts away from the big picture. Use of only weekly charts, although they clearly show the direction of prices, present higher risk and little opportunity for a good entry point. The obvious solution is to combine both charts into a program that uses each to its best advantage.