IBM is believed to have made the assumption that at most 50 to 100 computers would ever be needed to handle all of the world's data processing needs. Hence, they set-up up production, cost estimates and price estimates based on this "guesstimate". How ironic to think that probably any U. S. college's incoming freshman class now has more computing power in their dorm rooms.
How do I tie in the growth in "ultra" large firms with the sole proprietor/small employer? Is government as a monopoly finally outliving its usefulness? Is "Balkanization" really an attempt to get more local government? Government which is more responsive to the needs of the people?
I'm not sure whether it's Alan Greenspan (and the Fed) who is confused or the press and politicians. Then again maybe it's me.
With all the hand-wringing about six increases this year in the discount rate and/or the presumed target for the Fed funds rate I cannot recall anybody mentioning that most of the time the Fed adjusts its interest rates -- the aforementioned discount rate and Fed funds rate -- to reflect the reality of the short-term interest rate market. Yields in the short-term market have been significantly higher than what the Fed has been charging for money. Investors have not been anxious to purchase short-term U.S. government debt. To allow huge differentials to exist between what the market wants for a yield and what the Fed is "willing" to charge amounts to a subsidy for banks and others who can borrow at these preferred rates and purchase short-term debt. It has never been my understanding that banks were supposed to be the primary purchasers -- and beneficiaries -- of the short-term government debt market.
However, the Fed has slowed growth in the money supply. Depending on your definition of money and its measurement, they might be strangling the economy. Less supply with demand from a slowly growing economy suggests higher rates.
If there has been an effect, it has been on bond and equity prices. Using the long-term bond and futures markets, debt instruments are down by approximately 15%. Major stock market indices -- such as the Dow-Jones Industrial average and the S&P 500 -- are down about 5% from their highs. Other indices are down much more sharply -- the Dow-Jones Transportation average is down over 30%, the Dow-Jones Utility average is down over 25%, the Nasdaq is down, the Value Line index is down.
The stocks which have supported the indices have been much like the ones I have been discussing -- players in the "digital" revolution. Since June Apple (AAPL) is up, IBM is up, Motorola (MOT) is up, Microsoft (MSFT) is up, Intel (INTC) is marginally higher. These companies are bringing newer, better, faster, and cheaper technologies to market. They keep generating more business through lower prices and eventually cause significant changes in other products we purchase. As the year ends expect some signs of tax-loss selling; most of it is usually done by mid-December. Even in the bond market, although what may go on here will be "swaps" of losing positions for positions in bonds with similar characteristics. With tax rates up (and perhaps going lower), many will allow Uncle Sam to share in their losses, because he is so willing to take ever more of the profits.
This leads me to think that next year will start out well, probably end higher, and volatility will mirror this year's. There will be a number of 5-10% moves in the major indices with at least one spike up and, perhaps, one spike down. We are entering a time in the market cycle that politicians (and the Fed) find irresistible; the economy must get going. Higher interest costs have hurt smaller businesses -- this will change. Large corporations on average are trading at more reasonable price-earnings ratios than six months ago -- at least 15 percent below the average at market tops. If earnings continue to improve (and surprise, even), and if they can raise dividends, then their prices could rise 10 to 15%. These averages (DJIA and S&P 500) have traded in a narrow band for a long time, and they are now "fairly" priced; we need a sharp spike upward with a sense of giddiness and feelings that this bull market is completely different from all prior bull markets.
I would not be surprised if at least one Fed governor expresses a view that inflation worries have been overstated. This could be a boost to the smaller capitalization stocks which I think will recover most or all of their price declines of this year.
Long-term bonds do not excite me here. I think they will re-test their lows. If you know you would be a buyer with yields about 50 points higher, I would consider selling some 95 or 96 puts on the March 1995 US T-bond future. And if you are somewhat bearish I would also consider a bear spread -- buying March 99 puts and selling March 96 puts in a strict 1-to-1 ratio.
Would you trade one or two professional sports (for at least one season) for a better Christmas shopping season? A better overall economy? Let's see ... arrogant owners who "fix" prices and try to prevent "free trade" versus immature athletes with incredible physical talents who think they are entitled to a lifetime of leisure after "entertaining" us for several years -- as if the job market couldn't find something else for their skills (I'm assuming "whining" isn't a skill). And each side seems to think a "tax" might be part of the answer. Makes me wish they could both lose!
Why does the government try to "fix" markets, such as health care, by "fixing" prices? The largest (and most inefficient) monopoly in the country (if not the world) with a Justice Department which is ostensibly "anti-trust", and they are telling us that the "cure" is to mandate prices and choices!
Instead of the continuing "Balkanization" of our schools and universities perhaps we should re-unite most of our disciplines. Put Sociology back with Economics (most social activity is very much like traders making a market), and then put Economics back in with Politics, so perhaps our politicians (... at least political "science" students) will understand the economic consequences of their legislative proposals and the "lessons" they wish to teach the emerging economies.
And within Economics, place more emphasis on "micro" and markets, so the brains who understand the economy at large will begin to understand that every time they propose a law having economic consequences (which sounds to me like every law and every activity), then if you separate the legislated behavior from its costs or consequences, you will always end up with "deficits" -- or at least asking the middle class to subsidize someone else's notion of "wouldn't it be nice if we could do/pay for this".
The term "revenue neutral" (which was used with great frequency in several races in the recent election and sounds as if our elected leaders are equipped to make such decisions) has got to be one of the newest oxymorons. Especially when "neutrality" seems to discourage the more desirable behavior and encourage the less desirable behavior.
Do we have any idea on the limits of computing power possible in a personal computer (dare I say "hand-held")? A recent NY Times article cited a computer scientist: "Molecular computers ... are breathtakingly fast and efficient, and they have unheard of storage capacities. ... molecular computers can perform more than a trillion operations per second, which makes them a thousand times as fast as the fastest supercomputer. They are a billion times as energy efficient as conventional computers. And they can store information in a trillionth of the space required by ordinary computers."
An ad for Intel's Pentium processor was produced on a Mac -- a Quadra 950. Several Mac publications, including MacWorld, have reported Intel hired an outside group to oversee production of the ad. That outside group subcontracted with a well-known animation facility which owns no DOS or Windows machines, and due to business conditions had to use its Macs to meet their deadline.
Will the fact that the DJIA closed up 41.72 points on expiration Friday, December 16, 1994 finally disabuse the pundits who think the only big move possible on an options/futures expiration is down?
In the past six months there has been a quiet growth in the number of software packages available for the Newton. The price of the original Newton model 100 is now $199 through a number of dealers (which leads me to conclude there will be a new model within the next few months), and the newer model 110 is now bundled with several software packages while being priced about 5% lower than in June. I hear quiet murmurings that sales are beginning to get respectable. Unfortunately, John Sculley promoted it as if were something like a Cuisinart, when in reality, it should have been sold as if it were a "Vegematic" -- albeit one with a great deal of potential.
The PowerMac is also having a quiet boom in the software which has been developed and/or "compiled" to take advantage of its technology. Almost every observer forgets that the technology must appear in the marketplace (not just be "vaporware") before developers will actively look at it and think of the "neat" things they might be able to make it do.
And Apple's partner IBM seems to be doing things right. Two quarters (so far) with earnings better than expectations. Signing several more deals with Apple (which should give even more credibility to the PowerPC/PowerMac). Putting out their own system upgrade, OS2/"Warp" (with very clever commercials) -- which is getting very good reviews. In addition to being available at least six months before Windows95, Warp works well with only 4 MB of RAM. In contrast, the pre-release test versions of Windows95 need at least 8 MB to function.
Meanwhile Chicago/Windows95 looks more like 95.5 or 96. They are demonstrating it, but the roll-out date has just been pushed to late-summer-1995 or later. Nobody seems to point out that this operating system will have the same "flaw" the Mac OS has when running PC programs (MS-DOS and Windows) in emulation -- namely, these programs will run slower than they currently do, and there is nothing Microsoft or IBM or Intel or anybody can do about it; you will need a Pentium-based PC and most older programs will run as if on a 286-based machine (three generations of this technology ago). Additionally, nobody seems ready to step up and purchase the Pentium machines, and Intel seems to have hurt its reputation with the handling of the division error the Pentium is susceptible to making. SoftWindows (for the PowerMac) has been written by the same company, Insignia Solutions, which is developing a similar program for Windows95 -- with the full blessing, agreement, and cooperation of Microsoft; in 1995 expect SoftWindows to improve so that most programs will run as if on a 386-based machine.
There are many users (probably most) who do not need blazing speed and lots of extra features, but that is not how most buyers make decisions. Each user seems to find several nifty things he can appreciate doing which 98% of PC users would never consider (but 2% can be a great market share/saturation for a small developer). Yet speed remains an important consideration for fax-modems and CD-ROM drives.
With Motorola's entry into the personal computer market I can see the PowerPC chip easily reaching 30% of market sales. Does Apple holding near 12% -- within the next year every new Mac will have a PowerPC chip, IBM holding near 12% -- most of their new production will become PowerPC-based as the Pentium fiasco helps this happen more quickly, and Motorola getting 6% seem far-fetched? What if several others start cloning with PowerPC chips? In leading edge sales over the next several years I think it will gain significantly on the Pentium/P6 (next generation) chips from Intel.
The recent announcement by Intel of a problem in handling "rare" instances of real number division in which the answers are reported to be in error at the ninth decimal place brings to mind two completely different thoughts.
First, in the handling of the discovery of the error Intel erred -- especially if they learned of the error last summer. In a situation where they had just uncovered the mistake I would tolerate an early misstep as long as they got the treatment of their customers right, but they have had several months to consider actions, and IBM (one of their larger customers) got their response correct on the first try. And perhaps Intel has now gotten the right response; although I'm surprised by the rebound in the stock's price back to its levels before the announcement -- I think there are too many unknowns in how this incident affects their bottom line.
Second, the computer revolution could be making us more "innumerate". We are relying on these machines to do so much and to do it quickly and correctly, that we forget how rapidly a computer -- whether hardware or software -- can make mistakes. If current programs can execute one million arithmetic instructions in one second, and if an error can occur just "once in a billion" times, then we are talking about an error every 1000 seconds, or every 16 minutes and 40 seconds if the computer is constantly performing arithmetic while using a wide variety of algorithms. This estimate is extreme, but we sometimes lose sight of the fact that it might be sufficient only to know that it is December 1994 rather than the exact second, or that 2 + 2 is about 4. On the other hand, one in a billion may not be that rare if the expectation leads to several errors per day, and these errors cost large amounts of money to correct.
But where did Intel's estimate come from? In a recent column in the Chicago Tribune the following example was cited:
5,505,001 divided by 294,911
Every time you do this, you'll get this answer on your $4 calculator and virtually any other adding machine:
18.666651973.But try that same bit of arithmetic on [a] brand-new Pentium-based desktop computer and you'll get this:
18.66600092909.
An error in the fourth decimal place. In many real-world situations this is an intolerable error. And it is one which even an expert user will probably not notice until a subsequent answer does not look right.
Yet experts are surprised. It must be the location, because "rounding" errors occur all the time. One key to accurate programming is minimize the times numbers of significantly different magnitudes interact and to build in redundancy (i.e., a second method or "check" which "validates" the first method).
We have similar examples in the O.J. Simpson pre-trial testimonies. On the reliability of DNA testing we hear estimates of likelihood that samples did or did not come from the accused. These numbers are cited as if there is some certainty involved, yet I keep thinking of the description of statistics as "lies corrected by guesswork". In cases such as this, it should be easier and more accurate to say something did not happen than to say it did; if we cannot definitely say that an event did not occur, this does not mean it did occur. It makes me wonder about my concept of "reasonable doubt".
And in the worries about air safety. The most dangerous parts of a flight are the take-off and landing. "Commuter" airlines do these relatively more often per mile flown than the large carriers and often into smaller airports which may not have the economic or political clout to get the most sophisticated equipment. I am not surprised they have a "worse" safety record. Should we "require" expensive upgrades in aircraft when they may not significantly affect the safety of operations but may reduce the numbers of cities serviced and force more travel by less safe methods?
We see a number in print and think of it as a "certainty". Prices in the marketplace are only certain when the trade is made -- otherwise, they are only estimates of the current value of a product ... but the best ones we have at the given moment. One in a billion could be very costly; one in million could mean there are on average at least 12 people in metropolitan Los Angeles who match O.J.'s DNA -- which is significant if he's innocent and possibly irrelevant if he's guilty.
Prices going still lower for Macs, PCs, and all their peripherals!
Estimates from earlier this year (see Inside the PowerPC Revolution by Jeff Duntemann and Ron Pronk) show Motorola able to produce the basic 601 PowerPC chip for about $75; Intel's estimated Pentium cost is $450. A recent price (minimum order: 1000 chips) of the 100 megahertz (MHz) PowerPC 601+ chip is $399 from IBM and $439 from Motorola; the new PowerPC 604 running at 100 MHz is $549 from IBM and $599 from Motorola; the 100 MHz Pentium is $935. This PowerPC 604 chip is estimated to have 60% greater integer performance than and twice the floating-point number performance as a Pentium.
If Apple (and IBM and Motorola) can increase production, I think they might more aggressively price the PowerMacs. They are bundling them with modems and CD-ROM drives which are also tumbling in price. Modems with speeds which were prohibitively expensive 5 years ago are now priced near $100 and that includes fax send/receive capability. Faxing will fade in importance, because software exists and is improving which will allow you to pass the whole document to another user without any degradation (as in faxing and scanning) and the transmission times for these documents will be faster than for any facsimile. There are at least four programs -- from Adobe Systems (ADBE), Farallon Computing, and Apple among others -- I am aware of which allow a computer user to create a document and give it to someone who does not have the software used to produce that document and the recipient can view it on his own computer (and print it if desired); several of these allow for Mac to PC file creation (or vice versa).
I have seen a triple-speed CD-ROM for the Mac priced under $200. It was less than eighteen months ago that the $300 barrier for a double-speed CD-ROM was broken.
Hard drives are becoming smaller and cheaper. LaCie, the Mercedes/Rolls Royce of manufacturers, has a hard drive for the Mac with 730 MB of storage for under $400. This is lower than anything I've seen for any drive up to and somewhat beyond this size -- last year this would have been well over $1000. At larger sizes the cost approaches that of the equivalent storage on 3.5 inch diskettes.
Laser printers and color printers (all technologies) have again had significant price reductions with improved capabilities.
Where have you been invested for the last six months? If it's been in the Dow-Jones Industrial Average, you are up a little. From the close June 15 to the close December 15 the DJIA is up 49.08 points or 1.29% -- add half of the dividend yield of about 2.8% and you are over2.5% ahead (ignoring any commissions). The long-term US government bond market is down about 4.5 percent but if you adjust for interest income, you are down less than 1 percent. The numbers are similar for the S&P 500 average -- up 0.25 points or 0.05%. The same could be said of the DJ Utility Average which is down about 2.5% since mid-June. Both the bond market and utility stocks are down significantly -- 15 to 20 percent from their highs but most of this damage occurred in the first six months of 1994. In the last six months nearly all US indices are down less than 5 percent.
On the other hand, the DJ Transportation Average is down 12 percent since mid-June, and some market observers have estimated that as many as half of the NASDAQ stocks are down 40% or more from their highs and 60% are down 30%. This sounds like a "crash" to me!
And signs of year-end tax-selling have shown up. There is such an array of stocks that if you've got a NASDAQ favorite which has been beaten up this year and it's been making new lows in the last few weeks, I'd consider buying it (or even them) for the next month or two at least.
As far as bigger issues go, AT&T (symbol: T) has been making lows (low: 47 1/2; recent: 50 1/2), but the communications industry has such profit potential that I like it for at least the short-term. The insurance industry has been battered as Sears (S), for some reason I can't really fathom, has decided to let Allstate (ALL) (low: 22 5/8; recent 23 3/4) go after 60 years. With an additional 4 times the number of shares in ALL about to be transferred to current owners of Sears that is a large potential supply, but the stock is right near its lows, it has generally done very well in its investments, and the industry has always been cyclical.
Finally, I still like Apple (AAPL -- recent: 38 3/8), IBM (IBM -- 73 5/8) and Motorola (MOT -- 57 5/8) versus Intel (INTC -- 62 5/8) and Microsoft (MSFT -- 61 1/2). Motorola has been trading in a way which reminds me of Federal Express (FDX) shortly after it was first listed in the early 1980s. This stock has some big moves which can allow you to use options for short-term hedging or speculation, and I expect this to continue for some time -- increasing lows with frequent 10% or larger moves up and down.
I also believe we are coming into a strong period for stocks. As far as the political cycle is concerned the first of the two years prior to a Presidential election tends to be above average. With so many stocks significantly lower than early- to mid-1994 I think there could be some softness in bond prices -- perhaps testing the recent lows -- as funds get moved from 8% yields to stocks which could rally significantly in the next 6 to 12 months.
Is Orange County the tip of a huge iceberg of leveraged municipal debt? I doubt it, although I'm sure there are at least several other municipal bodies who have made a similar mistake on the direction of the market. The bond market is down about 18 percent from its highs which suggests to me that nearly every bond portfolio in the country is worth at least 10% less than it was early in the year! I have been under the impression that their portfolio is down about 20% which might mean they have just under-performed the market.
Again we see an example of at least one large investor not understanding that the leverage of derivatives pulls in at least two directions -- you can increase or decrease risk to almost any extent you like (I say almost, because to me the only time you have 0% risk is when you don't want it ... and you'll have it for a long time!).
I have to wonder who was advising these people. Where were the bankers? Who were the brokers? And did they take the other side of these trades? Did they "shop" this trade around? How anxious were the brokers to take the other side of the trade? Other than Orange County's threats to sue, even the lawyers have been quiet, but the politicians have been getting their sound bites.
Who was on the other side of these positions? If it turns out to be other customers, then perhaps we have an example of people who were using derivatives to hedge risk ... or speculate "intelligently" (i.e., if you make money doing it, you were obviously doing it intelligently).
The press in their haste to misunderstand a "debacle" tend to worsen the situation until our elected officials exacerbate the misunderstandings even more. At least Alan Greenspan had the guts to point out how much financial liquidity has grown since derivatives came out of their "dark ages" with the opening of the Chicago Board Options Exchange in 1973.
Major Microsoft products, such as Excel and Word, have been the product of a design philosophy crudely summarized as, "Finding out what features people want and programing them. We don't optimize or fine-tune unless absolutely necessary, since that takes too long and the chip makers will take care of our performance problems eventually. CPUs get faster and memory gets cheaper, and we want to get products out the door now."
What does this lead to? First, in the case of Windows95, the need for an extra 4 MB of RAM cost at least $100 in today's marketplace. If the hard disk space necessary to store the system is 25% larger than it could be (say 50 MB rather than 40 MB -- this is not that far out of line), you are adding at least $10 to the user's cost. Second, it leads to program updates which "fix" all the errors and "glitches" that manage to get into products rushed to market -- updates which customers pay for. This tendency for incredibly large programs is beginning to appear everywhere -- even Apple's new System 7.5 really needs more than 4 MB of RAM installed; although many users can do basic work with 4 MB, the ability to have a second or third program "open" is often desirable and almost always requires more than 4 MB.
Grain prices seem a little screwy.
We had a record year for corn and soybeans -- one year after floods which were thought to damage the whole industry. Prices of both are low but still more than 10% above the lowest lows of the past decade.
Soybean future prices are about 2.5 times the same month corn futures. This is at the low end of this ratio; the nutritive value of a bushel of soybeans is generally between 2.7 and perhaps 3 times that of corn. The ratio of their prices usually runs from 2.5 to 3.5. At these levels I think a lot of corn will be planted relative to beans, because if there is early season damage, most corn fields could be replanted with soybeans.
On the other hand, wheat is still fairly expensive with the futures being slightly inverted (Dec94 was and Mar95 wheat is higher than the later months). A few years ago it was impossible to find buyers at $2 per bushel after some record crops. I think last year's $4+ prices were an extreme in the other direction.
An interesting spread might be a variation on corn-wheat using long soybean calls against short out-of-the-money wheat calls and long out-of-the-money wheat puts. Something where the exposure is about twice as many bushels of wheat versus the soybeans would seem appropriate (unless you want to buy more puts) -- for example, buying 1 Jul 575 soybean call while selling 1 Jul 350 wheat call and 1 Jul 360 wheat call and buying 1 Jul 340 wheat put and buying 1 Jul 330 wheat put. Recently July beans have been near $5.85 a bushel and July wheat has been $3.45 a bushel.
Why did health care "reform" really fail?
We seek simplicity, yet we have a distrust of it. If we had a government-mandated plan, we would have given up choice. I doubt any of us really wants that. And a government plan would have just pulled costs out of the health care equation meaning incentives would no longer have existed to control many costs and improve many procedures.
Much of the complexity of modern world is the vast array of choices we can make in so many areas of our lives. I think this is why people claim to not understand derivatives, yet the basic ideas are concepts we have dealt with all our lives; it is the number of ways we can derive new choices.
Companies sometimes offer too many products which are only marginally different from their basic offerings; then they trim their "menu" but never eliminate choice. Just as markets fluctuate around a "true" value, many product lines fluctuate around the "correct" number of choices.
* A great word I first saw in WIRED! It's been around for a long time (I looked it up in the OED).
Greg Cramer
O.M.S.
P.O. Box 154
Winnetka, IL 60093-0154
(773) 334-3470 -- Voice/Fax
Date created: December 1994 Last modified: August 1, 1997 Copyright © 1994, 1997, Greg Cramer, O.M.S. Maintained by: Greg Cramer optionsms@earthlink.net