Jul

6

DOW 1000, from Al Corwin

July 6, 2010 |

I don't mean to be uncharitable to mathematicians because I would be in deep trouble without them, but I think this article does indicate why you have to take anything a mathematician says with a grain of salt. That is, the main concern they have is often solely whether or not they get the math right. The main concern should always be whether the math provides an accurate description of the situation. I can't judge the quality of Predhter's math, but I don't think we will see DOW 1000 even if we see a nuclear attack. Where are the parallels with the South Sea Islands bubble other than a few numbers chosen somewhat arbitrarily?

Rocky Humbert writes:

The Dow Jones Industrial Average hit a high of 380 on 8/30/1929 and a low of 41 on 7/8/32. That's a decline of roughly 90%, and it occurred both before the advent of nuclear weapons and less than 80 years ago. (Note that this period also included a 30+% decline in money supply and nominal GDP.)

So, Prechter's prediction– whether right or wrong– has recent precedent, and cannot be summarily dismissed.

Yet, if I were going to make an outlandish prediction for the purposes of garnering attention and selling newsletter subscriptions, it's that the S&P will decline by 60%, rally by 75%, decline by about 60% and rally by about 75% repeatedly over the next few years; and only subscribers of my newsletter will be able to nail the exact lows and highs. Unfortunately, a confiscatory capital gains tax rate will mean that my readers will have the exact same after-tax return as owners of T-bills.

Nick White writes:

I'm not the sharpest knife in the drawer by any means, but I just don't understand TA at all. Why would markets by necessity have to conform to some kind of relatively arbitrary pattern that someone has divined - especially when we well know that randomly generated time series have trends and patterns that can be discerned? It just seems like induction gone manic. All this stuff on fractals, waves, sequences, trends etc…it just seems like wishful thinking to me. There's no hard reason for any instrument to make its next step BECAUSE it's conforming to some silly sequence…there are block orders, agents, funds, indices and a million competing viewpoints that all bear on an instrument. I'm happy to be educated, should someone wish to explain it to me in a scientific vein, but I can't help but believe there are much more evident - though tough to quantify - elements that bear on the day to day movements of securities.

A small fable to illustrate my gripe, if I may: On t0, our hero, TA diviner (Mr. TA to you), sees some particularly juicy long setup appear in some retail stock (XYZ Retail)…. a real once in a lifetime thing according to his system…and yet, contemporaneously, we have Fund Manager (FM) - a standup guy with 12 years experience since his time at Wharton and HBS, but who's getting a bit bored with the industry and is going through a messy divorce. All is not lost though, as FM is, this very day, having lunch with a very attractive broker from Goldmorgan Sachsley (Ms. Lovely). At age 25 with two years broking experience and an art history major from Princeton (no offense), our lovely explains - over a very nice wagyu beef sirloin (grade 9) accompanied with a suitably expensive bottle of Pauillac to lubricate discussion - that she no longer likes XYZ. Indeed, she goes on to say, she and her friends don't like shopping there anymore; the hemlines are far too long this season and this is New York City, after all. Being very excited at this tidbit - and in an attempt to procure further lunches with said broker - FM decides to unload his 22.45% holding of the company right into the tiny buy orders of Mr. TA. TA man keeps buying on dips, but FM keeps unloading - after all, those broker rankings are tough to ascend and, besides, it's a good excuse to ring the lovely broker.

Now, a few days later, FM's mate - FM2 -knows FM "has the pulse of the industry" and asks him why XYZ Retail is down 5% this week when it was looking so good. FM2 mentions to FM that he's thinking of buying. Of course, as we noted, FM is a stand up guy and soberly reveals - just quietly, of course - his very grave concerns to FM2 about XYZ's prospects this quarter. His extensive, proprietary research has indicated that same store sales are looking grim. FM2, having never really had an original idea in his life (and having lucked his way into this gig through his uncle anyway), predictably keeps to his usual form and decides now is as good a time as any to short XYZ. He feels guilty for a moment, but then reminds himself that he is well behind for the year and could use an easy layup; those school fees certainly aren't getting any cheaper and the sting of his wife's expression on bonus day last year drives the chip ever deeper into his already strip-mined shoulder. As soon as he's placed his order, he does the ring-around to his other portfolio manager friends - giving out the warning of FM but without any of the stand-up disclaimers. XYZ falls further.

And so it goes for a few days…until Mr. TA - having resigned himself to a nice 18.47% loss on his promising setup (before margin) - finally sells his holding…. to a very smart, young, junior PM (Ms. Goodvalue) from a massive, arabic family office (actually, Goodvalue was a classmate of the lovely broker, though a different major. They'd been on the same varsity sports teams at Princeton, so they would acknowledge each other at parties in the City, but the lovely broker would always be slightly condescending to Goodvalue, while Goodvalue would assuage her subsequent insecurities (lovely *was* taller, after all) by telling herself that her honours degree was worth far more in the longrun than lovely's uber designer wardrobe and predictable future cosmetic augmentations). Goodvalue knows the company inside out, has done all her homework and cannot believe the discount that XYZ is currently trading at. She'd been waiting for a good opportunity to pick up XYZ and would have even begun buying some last week, if she hadn't been on vacation with her fiancee. In fact, she was only barely able to constrain her anger at both him and herself at the end of the vacation due to the missed opportunity. Now, she can't believe her luck. Goodvalue has years for this one to come good. She buys cautiously, with very deep pockets.

Meanwhile, Mr. TA has finished unloading his position. He chalks this one up to bad luck - after all, it looked too good to be true, and he should have known better than grabbing at such an ostentatious opportunity. He'll get it back next time; he always does. He's even considering going short now, the trend looks to have convincingly turned. He decides to short it on the open tomorrow.

The very next day, at 4:30pm the CEO of XYZ discloses a large purchase for his own account, as well as a contemporaneous buy-back; citing the extraordinary opportunity created by the last few days' indiscriminate selling.

Here ends the fable.

I part with this; there was a story in the news yesterday about a woman in Texas who just won the lottery for the fourth time…three of those times with tickets bought from the same store. Can't wait to buy her book and subscribe to her newsletter….she must know something.

Anatoly Veltman writes:

1. Markets have changed significantly since open-outcry execution was replaced by electronic, and algorithmic floodgates opened. Whoever claims that nothing changed, because human psychology has not changed (fear and greed, etc) - is clearly out of bounds instantly, as humans no longer place/cancel majority of orders.

2. Even pre-electronic era, Technical Analysis (TA) was hard to define. Most participation has always been technically driven, whether such was realized and admitted, or wasn't.

3. Clearly, there are no tools that create profits by themselves - or they'd become so widely replicated that they would stop being profitable. TA tools are useful only to those participants, who understand them in depth. Then you may know/feel when to use them and how to use them.

4. TA tools have always been suspect vis-a-vis individual stocks, especially those that are not extremely widely held, or of not extremely high market cap.

5. I'll now risk more controversial remark: the most important basis for history repeating itself has arguably been shattered in recent years. The moment the free-market system died - so did TA, largely.

Notwithstanding all of the above, I believe that opportunities can still be found - especially in the most over-extended, craziest market situations. It is precisely there - that the algorithms may be shut down, and individual experience and discretion may take over. At the proverbial "back to basics" time!

 

Bruno Ombreux writes:

How coincidental!

I have just started reading "Bayesian Forecasting and Dynamic Models" from Harrison and White. I quote from Chapter 1: "A Time Series model is generally a confession of ignorance".

It hit home because I had already formulated it myself, when I was 16 years old and was first taught the bell curve: "a probability model is an admission of ignorance".

It is really easy to get drawn into complicated models and the maths, easily forgetting that the reason why we are using models is because we don't have a clue about the real world to start with. The fact that the model is probabilistic instead of deterministic means that we are even more clueless than clueless. We are the epitome of cluelessness.

I'll take a real life example from trading crack spreads.

Crack spreads are a proxy for refining margins. If you own refineries, really, you don't need a statistical model. To start with, you know your real refining margins, not the proxy. Let's say the cracks are going down, way down. As you own the refies you'll know when you start shutting down units, curtailing supply then pushing margins back up with a delay. No need to resort to statistical models, you have quasi deterministic knowledge of when to buy the proxy. That works if you are a refiner. Guess why trading houses like Dreyfus or Vitol bought refineries in the 1990s?

Now let's imagine you are a sophisticated, I insist, sophisticated, hedge fund trader. You don't have a clue about the physical world, but obviously, from the behavior I described above, crack spreads are mean reverting. You can do your statistical study and figure out the spread is 4 sigma away from mean. You buy it. You are at a lower level in the food chain, because the refiner gets first dibs. He is Lion King, your are a hyena in her pack. He'll know before you, because he is the one who will *cause* the mean reversion. And more importantly, he'll know when not to buy a 4 sigma because it will move to 5. You used statistics because you are clueless. But you can still make money, there are leftovers. Don't be too quick to dismiss price patterns, mate.

Let's move to the third level in the food chain. These are the people, wholesale and retail, that are not even using statistics, but stuff like non-scientific TA, eg Elliot Wave or various indicators, or naive fundamental analysis, by naive I mean without the data available to a refiner, and with poor logic, that is 99% of research provided by banks. These guys don't stand a chance really. They get third dibs and that means there is really no food left on the table for them. They are the food on which lions and hyenas prey. They are the noise traders of academic fame.

Rocky Humbert writes:

Mr. Bruno's wisdom misses one key point: The refinery business has been, over long periods of time, a lousy business. If you catch the cycle right, you might make a few bucks, but over the entire cycle, domestic refining has provided a lousy return on equity, as well as exposing the company to counterparty risks, environmental and other liabilities etc. etc.. Refining is not quite as bad as airlines, but it's bad.

One can therefore surmise that integrated oil companies which remain in the refining business do so for strategic reasons — and their other business subsidize their refining businesses. If one accepts this proposition — that refiners behave in non-profit-maximizing fashion, economic theory dictates that there are actually profitable opportunities for speculators, despite Bruno's evident derision for market participants who are lower on the food chain. The most recent memorable example of this was the June 2009 crack spread which traded at a negative value in the fall of 2008, and which was a textbook illustration of how speculators can profit at the expense of refiners

See this.

[The Chair may recall that he and I corresponded about the evident mis-pricing at that time.]

I don't quarrel with Bruno's comments regarding blindly following statistics. However, there are ample opportunities for speculators/investors to profit when intelligent fundamental analysis, risk management, and technical analysis are sensibly applied. The definition of "intelligent" and "sensible" are left as an exercise for the reader — but they have precious little to do with statistics.

Bruno replies: 

Refining is a lousy business. It is so lousy that many majors are trying to move away from it and have been selling or cocooning refineries. All the money is made upstream, in crude oil production.

This doesn't mean that the refining business is cross-subsidized. Integrated oil companies are split into business units with a clear separation between upstream (exploration and production), midstream (trading and transport) and downstream (refining and marketing). Beside these major subdivisions, you have further splits, for instance a business unit for marketing jet fuel to airlines, another one for selling bunker fuel to shipping companies, another one to operate a local gas station network…

The downstream refining business sucks. It is feast and famine. It has low profitability. But this doesn't mean that within this low profitability environment, it is not seeking to maximize profit or minimize losses.

I have worked as an oil trader for an oil major. So I have seen it from the inside. I can tell you I have never seen something as profit maximizing as a refinery. Low profits bring extreme profit seeking behavior. Fat profits on the contrary bring laziness. I am sure a company like MSFT is getting lazy with its crazy operating margins and piles of cash. A refinery is not.

We are also talking about the biggest companies in the world here, the oil majors. Don't think for a second that they don't fight tooth and nail for every little cent of profit.

Rocky Humbert writes:

I can accept the statement that "high profits attract new competitors," but given the choice between investing in the stock of Microsoft (12x p/e; 38% RoE) and VLO (17x p/e; -13% RoE), I'll let you buy Valero and I'll pick Microsoft. Or, I'll buy Coca Cola (15x p/e; 30% RoE) and you can buy TSO (infinite p/e; -5% RoE) if you prefer.

Or to paraphrase a famous speculator, "Nobody ever went broke making a profit."

Al Corwin adds;

I don't think the problem at Microsoft is laziness. I've worked with and around them for over twenty years, and the vast majority of people there still work fanatically hard.

My belief is that Microsoft has gotten to the stage where their own success gets in their way. If your project attacks the Windows/Office economic powerhouse, it will die in internal politics, or at the very least be put on the back burner. The Windows/Office group has trashed Web Office, cloud computing, and mobile phone development efforts again and again. The Windows group has quashed everything 3except the X-Box.

This is why Apple is different. Apple is trying to make the iPhone and the iPad the only computer most of their customers will ever need. If that trashes their Mac sales (which it hasn't, for sure), that's the cost of progress. The Mac group can contribute, but it is not allowed to get in the way.

Google also would love to make the mobile phone all the computer you would ever need, and other competitors have the same goal. They are there already for many young people, and they won't be long delivering something like that for us more old fashioned types. I've already seen a cell phone that had a projector screen and a plug-in keyboard as well as more horsepower than my main development machine had seven years ago.

There's a thousand other problems at Microsoft, but almost all are related to not wanting to trash their own business. They can't go seriously after the young developers because they make so much money off corporate development. These are hard problems to solve, but great problems to have. The challenge of making too much money to think straight would be nice for a change.

Bruno Ombreux writes:

Rocky,

The way I would generalize my statement would be:

"Enterprises with poor profitability are trying harder to maximize profits (or minimize losses), than companies with high profitability and good growth prospects."

The profit-seeking behavior of refiners was our subject of discussion. This is why I would limit my generalization to this point. And let's be clear, I am talking about professionally managed companies, like the refiners, not about Mom and Pop corner shops.

This profit-seeking behavior means they will try to maximize refining margins, and therefore affect the proxy: crack spreads.

This was the discussion. In fact, my generalization is irrelevant to the discussion, because all we need is refiners trying to maximize profits (or minimize losses), independently of highly profitable companies behavior.

Now, coming back to my generalization, which is interesting even if it is irrelevant to the discussion about crack spreads, I can only speak from personal experience. I worked in several big companies, in good times and bad times. It has always been the same. Always. When the times are good, we get build cushy offices, add layers of management, promote yes-men, multiply meetings, become political and arrogant. When the times are bad, we cut costs, get rid of the layers of management, replace the yes-men by obnoxious goal-driven achievers that were previously passed up for promotion, replace the meetings by kicks into various corporate asses to get them on the road doing actual things instead of sleeping in meetings, get rid of the politics and learn modesty.

Stefan Jovanovich writes:

The best way to achieve the state of grace that Bruno describes ("when times are bad, etc.") is to never let the retained earnings stay in the business. The seeming anomaly of 19th century finance (stocks had larger payouts than bonds) was based on this principle. So was the Mars family's attitude about office furnishings.

Rocky Humbert replies:

Bruno: If I were to paraphrase and generalize your statement, it would be:

"Enterprises with poor profitability and no growth prospects can recruit and retain smarter, harder-working and more effective employees and managers than companies with high profitability and good growth prospects."

I challenge you to provide any evidence of your supposition that employees and managers in "lousy" industries are more effective than employees in "good" industries. I suspect that you will find exactly the opposite– that is, the GM parking lot empties a lot earlier than the Microsoft parking lot. 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search