David Nichols Morning Report
Enviado: 20/12/2002 15:08FRIDAY a.m.
December 20, 2002
They don't make fun of weather forecasters any more
by David Nichols
My recent piece taking umbrage with the "consumer is dead" dogma set off a lively debate, as I fully expected.
If you're telling me that the stock market has serious problems because households are over-leveraged -- and this is the main point many are making -- then okay, you just may be right. But then again, you just may be wrong.
When I look at this next chart (courtesy of Adam), which shows household debt payments compared to personal income over the past 20+ years, it just doesn't look that bad to me. But you can make up your own mind.
The real point -- and the one I was sneakily trying to back into -- is that trying to divine the chaotic, infinitely complex future movements of the stock market from a data point about consumer debt is the equivalent of, say, trying to divine the future weather patterns on the U.S. east coast from the measurement of a specific sea surface temperature in the Gulf of Alaska.
They are both impossible tasks. The outcome is too far downstream from the initial data point, subject to too many other factors.
The only way to accurately predict the weather is to collect as much fine-grained data as possible about local atmospheric conditions, and to use computers to model future behavior based on probability. (Interesting side-note: the most powerful super computer in the world is being developed in Japan to model weather). The market is the exact same type of chaotic system as the weather. Only our data-gathering facilities about the causative agents of change for the market aren't as good as for the weather.
The main causative agent of change for the market in the short-term is investor sentiment. Someday, someone will figure out a way to model the chaotic patterns of the market in a highly accurate way, based on sensitive readings of initial conditions, and accurate modeling of future behavior.
Of course, just like the weather, the models will become less accurate over longer time periods. We don't give much credence to the 2-week or 2-month weather forecasts, but the 3 to 5 day weather forecasts are now extremely accurate. Weather forecasters are no longer the butt of jokes, as they were 10 years ago. Data gathering and computer modeling progressed to the point where the forecasts have become highly accurate, even though the weather is an "unpredictable" chaotic, dynamic system.
There is no reason why 3 to 5 day market forecasts cannot also be extremely accurate. I'll just say -- intriguingly -- that we're working towards this end.
There aren't many that are even going down the right path. I just don't think it's possible to derive any sort of accurate predictive model from fundamental data points. Most people think this is where the answers lie, but I personally feel it's a misallocation of brainpower. And I misallocated plenty myself before reaching this conclusion.
Ultimately there is very little traceable cause and effect from the fundamental data points we can see and quantify. The reality is that people end up gravitating to the fundamental news flow that fits in best with their pre-conceived notions about the world and the markets. You have to be aware that your market prism is colored by your own particular filters and opinions. We all have our own "rose-colored glasses" when it comes to the markets.
Your viewpoint may turn out to be right, or it may be wrong. But pre-conceived notions about the market will likely only end up costing you money. We've all learned the hard way that it's never a good idea to dig in your heels too much about a fundamental market opinion. It's better to stay as objective and flexible as possible.
Sentiment Dashboard
by Adam Oliensis
SENTIMENT TANK: The tank got a rush of negative sentiment fuel on Thursday rising from 60% to 66% full as the benchmark indices broke down out of their brief consolidations. If the 71% level doesn't act as a ceiling we could be looking at the tank near 85% fairly soon.
SHORT-TERM GAUGE: The Hourly gauge had backed off a bit in its down-cycle reading but continued in its decline phase on Tuesday. It's getting fairly advanced in its progression, so with Expiration on Friday and a host of chaotic forces at work, it wouldn't be surprising to see a pause that refreshes, and some uptick in price. Why? Positions will get squared away for Expiration and the coming holidays. Right now short-term short positions have more profits than short-term long positions. In the absence of negative shocks, squaring those shorts could add a hint of buying pressure.
MID-TERM GAUGE: The mid-term momentum of sentiment continued on down, dropping into the decline phase on this gauge from 42 to 48. Our Confidence Diffusion Index (CDI) picked up a couple points to 3 (on a 0-7 scale with 7 as maximum confidence). We saw acceleration within the decline phase, a break UP in negative sentiment and DOWN in price toward our next SPX target of 872-874. Note: Apologies for yesterday's inversion in which I called 42 by its "doppelganger" name, "58." The gauge showed a correct reading, it was my description that was inverted.
LONG-TERM GAUGE: The weekly gauge flickered from neutral just slightly into a decline phase, dropping to a reading of about 15.5. The CDI moved to a BEARISH 1. It's a very preliminary and tentative weekly signal.
December 20, 2002
They don't make fun of weather forecasters any more
by David Nichols
My recent piece taking umbrage with the "consumer is dead" dogma set off a lively debate, as I fully expected.
If you're telling me that the stock market has serious problems because households are over-leveraged -- and this is the main point many are making -- then okay, you just may be right. But then again, you just may be wrong.
When I look at this next chart (courtesy of Adam), which shows household debt payments compared to personal income over the past 20+ years, it just doesn't look that bad to me. But you can make up your own mind.
The real point -- and the one I was sneakily trying to back into -- is that trying to divine the chaotic, infinitely complex future movements of the stock market from a data point about consumer debt is the equivalent of, say, trying to divine the future weather patterns on the U.S. east coast from the measurement of a specific sea surface temperature in the Gulf of Alaska.
They are both impossible tasks. The outcome is too far downstream from the initial data point, subject to too many other factors.
The only way to accurately predict the weather is to collect as much fine-grained data as possible about local atmospheric conditions, and to use computers to model future behavior based on probability. (Interesting side-note: the most powerful super computer in the world is being developed in Japan to model weather). The market is the exact same type of chaotic system as the weather. Only our data-gathering facilities about the causative agents of change for the market aren't as good as for the weather.
The main causative agent of change for the market in the short-term is investor sentiment. Someday, someone will figure out a way to model the chaotic patterns of the market in a highly accurate way, based on sensitive readings of initial conditions, and accurate modeling of future behavior.
Of course, just like the weather, the models will become less accurate over longer time periods. We don't give much credence to the 2-week or 2-month weather forecasts, but the 3 to 5 day weather forecasts are now extremely accurate. Weather forecasters are no longer the butt of jokes, as they were 10 years ago. Data gathering and computer modeling progressed to the point where the forecasts have become highly accurate, even though the weather is an "unpredictable" chaotic, dynamic system.
There is no reason why 3 to 5 day market forecasts cannot also be extremely accurate. I'll just say -- intriguingly -- that we're working towards this end.
There aren't many that are even going down the right path. I just don't think it's possible to derive any sort of accurate predictive model from fundamental data points. Most people think this is where the answers lie, but I personally feel it's a misallocation of brainpower. And I misallocated plenty myself before reaching this conclusion.
Ultimately there is very little traceable cause and effect from the fundamental data points we can see and quantify. The reality is that people end up gravitating to the fundamental news flow that fits in best with their pre-conceived notions about the world and the markets. You have to be aware that your market prism is colored by your own particular filters and opinions. We all have our own "rose-colored glasses" when it comes to the markets.
Your viewpoint may turn out to be right, or it may be wrong. But pre-conceived notions about the market will likely only end up costing you money. We've all learned the hard way that it's never a good idea to dig in your heels too much about a fundamental market opinion. It's better to stay as objective and flexible as possible.
Sentiment Dashboard
by Adam Oliensis
SENTIMENT TANK: The tank got a rush of negative sentiment fuel on Thursday rising from 60% to 66% full as the benchmark indices broke down out of their brief consolidations. If the 71% level doesn't act as a ceiling we could be looking at the tank near 85% fairly soon.
SHORT-TERM GAUGE: The Hourly gauge had backed off a bit in its down-cycle reading but continued in its decline phase on Tuesday. It's getting fairly advanced in its progression, so with Expiration on Friday and a host of chaotic forces at work, it wouldn't be surprising to see a pause that refreshes, and some uptick in price. Why? Positions will get squared away for Expiration and the coming holidays. Right now short-term short positions have more profits than short-term long positions. In the absence of negative shocks, squaring those shorts could add a hint of buying pressure.
MID-TERM GAUGE: The mid-term momentum of sentiment continued on down, dropping into the decline phase on this gauge from 42 to 48. Our Confidence Diffusion Index (CDI) picked up a couple points to 3 (on a 0-7 scale with 7 as maximum confidence). We saw acceleration within the decline phase, a break UP in negative sentiment and DOWN in price toward our next SPX target of 872-874. Note: Apologies for yesterday's inversion in which I called 42 by its "doppelganger" name, "58." The gauge showed a correct reading, it was my description that was inverted.
LONG-TERM GAUGE: The weekly gauge flickered from neutral just slightly into a decline phase, dropping to a reading of about 15.5. The CDI moved to a BEARISH 1. It's a very preliminary and tentative weekly signal.