Date: June 5, 2002
My story in March about the market turned out to be dead wrong..., still licking the wounds of that one. On the other hand I still believe that there are basic market forces at work that will surprise even the most jaded of analysts in the last half of this year.
Let me explain.
1.) The Current Mentality
The semiconductor merchant market has been under siege for nearly three years.
Their investments in new technology have been cut to the point of just keeping
current production fabs in working order while development of the newer (300
millimeter) facilities are still in the proving stage and the impact of volume
production won't begin to be felt until late 2003 at the earliest. Older quarter
micron fabs are now being shuttered or moved to less performance demanding
applications while 175 nanometer fabs reaching their maximum utilization rates.
The ability to start reinvestment into new fabs will take a minimum of three to
five quarters of strong profits before it is fashionable to think of this as a
communications item to the financial community as a competitive stance strategy.
Once this process begins vendors will have to wait in line as the equipment
suppliers will become overwhelmed.
The market consolidation of memory suppliers is but one outward sign of that
discussed above. Collectively they are looking to spread their business (read
risk) into other areas which have less exposure to the financial gyrations of
the memory sector without adding any additional costs (rabbit and hat trick).
Such application specific devices such as Networking, Low Power, CAM and 1T SRAM
are interesting as they represent a very small percentage as compared to the
commodity memory space but are important to producers as their potential to earn
on a greater margin offsets the smaller production quantities. An important
example of this area is the ability of the Graphics SDRAM which has remained in
relatively strong demand while maintaining decent margins all during the current
downturn..., although seemingly attractive it must be also said that these
devices are much harder to produce as the performance requirements stress the
process corners and challenge the packaging backend.
In summary the semiconductor merchant memory market is now experiencing the
worst business climate in its entire history. Investment in new plants will not
reasonably freshen in pace until the average utilization rates of current
facilities reside above 80% for a minimum of 3 to 5 quarters.
2.) Timing
What turns a market around? Most often stated in textbooks is the concept that
markets turn around when management feels comfortable with the level of risk
associated with current and future business prospects. One problem with this
statement is that it occurs after the market has turned. Markets turn when
individuals take calculated risks with good timing. A simple fact is that for
markets to recover they must have something to recover from..., the deeper the
red period, the higher the black period (Yin & Yang of Markets). It is common
knowledge that the Semi's certainly have something to recover from. The question
is when and by what dynamic?
Demand and Supply are most often touted as the ruling factors in market
dynamics. The following table clarifies this graphically:
| demand high | demand low | |
| supply high | = | low |
| supply low | high | = |
The memory merchants have been in the three-quarter market for nearly three years {= , low and =} where costs are fully constrained by oversupply conditions. This graph has another dimension which I believe lends itself to tenets of Catastrophe Theory in which the graph becomes three dimensional..., a diagonal line running through the "="s is the top view of a fold in the graph fabric in which the "high" datum is low point (and conversely the {low/low, =} is the high point). Imagine a marble placed on the data field, it will roll toward and settle in the low portion of the graph {low/high, high}. Technology driven markets tend to follow this function or cease to exist because they are not interesting, low or not self-sustaining in profit. The market space names are as follows:
| low/low | Null or Abandon |
| high/low | Marginal |
| high/high | Sustain |
| low/high | Growth |
This model is used to identify the candidate processes under examination.
The ongoing consolidation process has worked its magic in the Null/Abandon and
Marginal spaces..., the number of producers are now greatly reduced. The
Sustain/Growth space could be renamed the Consolidation space wherein the
magnitude of scale amongst few producers achieves an effective oligopoly or
protective covenant from further entry by competitors. The effect is price
stability with assured and stable supply. Currently the market is operating in
the Sustain space. The question remains, when will the market transition to the
Growth space?
Investment cycles tend to follow the curve that our marble takes on the datum
space..., Abandoned or Null and Marginal are excellent analogues as are Sustain
and Growth. Investment timing lags in phase the market health, overshooting the
growth while undershooting the marginal. The Semi's have had several false
starts over the last several years which has led to most being rather shy about
voicing any bullish support of future growth prospects. The current protective
scheme to keep CEO's out of shareholder lawsuits relies on the return of the
corporate IT buying. Reports over the first and second quarter provide fairly
solid evidence that the recession experienced in the U.S. has come to an end and
that recovery is now progressing apace. Europe and Asia are also showing signs
of recovery.
The setbacks over the last several years have dealt a severe blow to the Semi's.
Beginning with the dot.coms followed in close order by telecoms, energy brokers
and the balance sheet brokers. No one is currently venturing a guess when
telecomm will recover (late 2002?). The dotcom or internet is now undergoing a
new incarnation, entry into the established business infrastructure as an
integral part and will be many times larger than either the telephone or the
adaptation of the computer into business practice and is expected to be the
center of productivity gains for many years to come. The market is at a
transition point with future growth based on sustainable demand not on consumer
nuances.
The demand is just beginning to build. It is hard to see against the background
of supply filling the cracks of depleted inventories currently but it is there.
The conventional wisdom suggests that IT spending will begin in the second half
of this year..., data suggests that this has already begun. The issue works
itself down into the supply side of the Semi's..., when will allocation become
the byword as supply dips far below demand for the current generation? The
switch to the 256 MBit DDR SDRAM as the full commodity memory will further parse
the marginal suppliers leaving those remaining in a very strong earnings
position. What does this mean to the demand side of the equation? It means that
tying down long term supply contracts are absolutely a must. The demand side
will show as a very firm indicator of this in the next several weeks as second
half supply contracts have been negotiated or are under negotiation currently.
If the price stays nearly equal or up will indicate an expectation of tight
supply in the second half..., conversely, if it is down, then there is
confidence that the supply will remain adequate. This is a collective bet so its
outcome will be extremely interesting.
Merchant producers have a certain amount of unused capacity available for a
pickup in demand but it is nearing a crossover point in which demand loading is
for the newer fabs which are near to full capacity. Crossing this point creates
a roving shortage wherein supply is delivered through more costly upside
arrangements. Finally the roving shortage evolves into a full blown allocation
shortage with attendant price increases. The effect snowballs with double and
triple ordering from the demand side. Concurrently the IT space spending habit
changes from "wait and see" to "better place the order" to insure that projects
come in on time. Will this occur in the second half? Suppliers are hinting at
tight supplies in the third quarter but they have done this several times before
with no effect. It may be that the demand side has gotten used to the suppliers
crying wolf and are just ignoring them..., or it could be that the earlier false
starts were just drill in preparation for what may prove to be the mother of all
shortages.
===================================
Pssst! We've updated our Shopping Page.
===================================