| Barron's
The Trader: Late-Summer Bounce Continues
(From BARRON'S)
Being late is the same thing as being wrong when it comes to stocks. That
unforgiving truth is a source of great anxiety right now for investors and
traders as they try to figure out what the market knows and when it knew it.
In a week when Wall Street returned to work in earnest, observers across the
sentiment spectrum could find reason to wonder whether the market had already
factored in their expectations while they were on the beach.
An investor
convincedthat oil prices have seen their highs and the election
result will
be market-friendly might similarly wonder if the latest four-week
bounce means
the market already figured all that out.
Another who's certain that the coming
corporate earnings season will
disappoint might ask whether stocks' weakness
in July and early August amid
strong profits had already digested this prospect.
Last week's activity wasn't quite dispositive on these questions, with a
jumble of encouraging and cautionary signals vying for attention.
The Dow
Jones Industrial Average gained 52 points, or 0.5%, to 10,313. The Dow
underperformed
the broader market largely due to the drag of Coca-Cola, down
5.5%, and Alcoa,
off 6.5%, both on reduced profit expectations.
The Standard & Poor's
500 gained 10, or 0.9%, to reach 1023, a 10-week high.
The Nasdaq surged
49, or 2.7%, to 1844, as a powerful upward reversal in the severely underwater
semiconductor group carried most technology shares higher.
The rebound off
the Aug. 13 low, which preceded a peak in alarmism about oil
prices, has now
carried the S&P 500 to a 6% gain in four weeks. For an index
that has
occupied the range of roughly 1060 to 1160 since mid-December, that qualifies
as a significant move.
A comparison with the two prior rallies off interim
low points (set in March and May) offers some perspective on where the current
bounce ranks.
The rally from the March low lasted two weeks and amounted
to a 5.8% move. The
next bounce took six weeks and covered 6.5%. That places
the current move right
around the average magnitude of the prior two and right
at the average duration.
And, notably, each of these rallies has topped out
at a lower point than the
prior one.
But of course, the moment the
market starts to appear predictable, it
rediscovers its tendency to confound.
Aside from the record of prior moves, the available evidence allows for
plausible arguments that both bulls and bears might pursue as part of internal
dialogues as they try to divine what's already discounted in stock prices.
The case for more upside would point largely to the recent market action
itself.
Last week's gain could pass for resilience in the face of nasty terrorism talk
and some worrisome corporate news.
Semi stocks -- which play on the id of
investors -- jumped 7% following a
wretched 30% decline, in the face of a
belated brokerage-house downgrade of the
sector Tuesday and a bad-but-not-terrible
outlook from Texas Instruments. Nokia
added upside oomph to the sector by
raising its sales targets after having
lowered them serially this year.
The Dow Transportation Average ended Friday on a new 52-week high, usually a
positive economic smoke signal. Bank stocks never joined the broad market at a
new low and have remained firm. There was some trader excitement last week, too,
that the S&P surmounted its 200-day moving average.
Less concretely,
optimistic investors were looking to play a political
perception game, front
running an anticipated electoral rally based on the idea that the market at least
believes it prefers a George W. Bush win.
The negative stance rests largely
on what may well become a sloppy
third-quarter earnings season. The tenor
of earnings-forecast revisions has
worsened. Yet it's still not clear that
the downside surprises in broad economic
measures and the upside shock of
strong oil prices through most of the quarter
are yet reflected in the numbers.
The hurricanes and balky consumer-spending patterns have also raised the
stakes for companies to recoup business in the final three weeks of September.
That very same surge in semi stocks, in fact, could be spun as a cautionary
sign, indicating a stubborn instinct to treat hope as an investment strategy and
to buy only in fear of missing a rally.
The above-mentioned anxiety of professional
investors trying to out-anticipate
the market is, no doubt, intensified by
their generally disappointing
performance this year.
Large-cap "core"
stock funds -- the category with by far the most assets and
the closest in
style to the S&P 500 -- is trailing the 1.33% year-to-date return
of S&P
index funds by more than two percentage points. Hedge funds as a group
are
up in the low-single digits this year.
Their managers face the prospect of
not getting paid this year, forcing them
to buy puppies rather than ponies
for their kids' birthdays. The prospect of a
few thousand hedge funds watching
the year tick away and operating in their
two-minute offense makes the job
of handicapping the market that much harder.
For nonprofessional investors
not held to artificial calendar-year performance
benchmarks, the story remains
largely as it has all year. Stocks have gone
nowhere and thereby gotten less
expensive without being demonstrably or broadly
cheap. Dividends from mature
companies are more plentiful but only marginally
more appreciated than in
years past, offering a head start for long-term gains.
And the math governing
equity returns still suggests that they'll be modest
for several years to
come -- meaning any forceful rally yearend may well borrow
from what 2005
might offer.
Scrutinizing Models
It might as well be
taped inside the lid of every market strategist's briefcase: The theory of the
Fed model, always handy when better reasons to buy stocks are elusive.
This
intuitive stock-versus-bond valuation model is a staple of market commentary,
having the twin virtues for stock salesmen of utter simplicity and a tendency
to make stocks look like bargains.
The model compares the 10-year Treasury
note yield to the inverse of the S&P 500's price-earnings ratio. "Fair
value" is supposed to reside at the point
where the yields are equal.
Last week's dip in the yield below 4.2% was
accompanied by an increase in
such chatter, showing stocks to be a yawning 40% undervalued.
The Federal
Reserve noted this relationship in the 1990s and inadvertently
conferred legitimacy
on a simple algebraic relationship that launched a thousand
investment pitches.
Barron's over the years has helped to propagate it.
Yet an overdue measure
of critical scrutiny of the model's logic and utility
is now being applied
by some investment professionals. The work of AQR Capital's
Cliff Asness in
debunking any theoretical substantiation for the relationship
has been detailed
here occasionally in the past.
It shows that, while low interest rates
(and inflation) coincide with periods
of high P/E multiples, they don't
justify them or predict good future returns
for stocks. Absolute trailing
P/Es are far better gauges of value.
Some Wall Street eminences have taken
up the issue. Morgan Stanley's Byron
Wien recently wrote an admirable and
thoughtful essay examining his own
long-employed stock-bond model in light
of Asness' work.
And just last week Smith Barney's Tobias Levkovich questioned
the Fed Model's
usefulness as a valuation and timing tool. He notes that in
the 1981-1996 period
covered by the Fed's original study, the statistical
measure of the model's
predictive ability (called R-squared) was reasonably
strong, at 0.77 on a scale
of zero to one.
But since 1960 to now, this
measure came in at 0.5. Since 1997, the model has
been half as good as a coin
toss, at 0.24. Rates, obviously, are but one
ingredient that make stocks taste
better or worse.
This shouldn't be shocking given that bond yields have fallen
from 6.5% in
early 2000 to the current 4.2%, even as P/E multiples have contracted
significantly.
It could be, in fact, that the only days of the Fed model's
practical value
were during the erstwhile two-decade bull market in stocks
and bonds. This
period saw virtually uninterrupted disinflation, falling interest
rates and
stock-multiple expansion.
The Fed model, perhaps, worked as
a means of telling which outer leg in this
three-legged race had gotten too
far behind the other and was due to catch up.
All this doesn't mean the market
can't rally. But if it does, it won't be
because the bond market told stocks
they were cheap.
Tale of Two Middlemen
To thrive as
one of the stock market's wholesale middlemen, it's necessary to
be huge,
smart or both. Executing stock trades is a fragmented and exceedingly
competitive
business, with prices charged to trade a share often reaching
fractions of
a penny.
Two of the smaller firms trying to thrive via smarts and servers
are Archipelago Holdings and Knight Trading.
Archipelago, a fast-growing
electronic stock exchange, recently debuted as a
public company and has quickly
attracted the fancy of investors. Knight, a
calloused veteran of the 'Nineties
boom, has been largely neglected by the buy
side.
Born in 1997 and backed
by several Wall Street firms, Archipelago is technically a stock exchange and
executes some 25% of all order volume in Nasdaq-listed stocks. The firm was
the first "alternative trading system" to link all major trading venues
on one platform. It completes about 85% of all trades by "matching"
orders internally between investors in its system, sending the rest elsewhere
for completion.
The company went public Aug. 12, pricing its shares below
its expected range
at $11.50. Since then, firmer markets and delayed investor
enthusiasm have
driven the stock up 32% to 15.
Profitable since 2003,
Archipelago is on pace to earn just under $1 a share
for 2004. Rich Repetto,
analyst at Sandler O'Neill, last week launched coverage
of the stock with
a 2005 earnings forecast of $1.11.
With $3.37 a share in cash, the Archipelago business, ex-cash, is valued at
less than 11-times forecast 2005 earnings. That's comparable to the valuations
of similar operations such as Instinet's electronic trading network (ECN), says
Repetto. He deems the stock a Hold.
Now that the stock has run up so far,
so fast, the looming unanswered
questions involve Archipelago's ability to
navigate the brutal economics of the
execution business and potential new
regulatory rules.
Transaction fees represented $251 million of Archipelago's
$275 million in
first-half revenue, with $24 million coming from sales of
quote data and a
negligible amount from fees paid by companies to list their
shares on the
exchange.
Archipelago received an average transaction fee
of 0.35 cents per share from
investors using its system to trade by accepting
posted bids and offers.
The SEC's current project reviewing market structure,
called Regulation NMS,
includes proposals to cap such transaction fees at
0.1 cents per share.
Another concern: Archipelago's minority owners -- including
brokers Goldman
Sachs, CSFB and Merrill Lynch -- accounted for 40% of its
volume. Those firms
sold shares in the IPO and could reduce their commitment
further.
And Archipelago is also intent on winning more outright stock listings
from
companies who would otherwise pay to be listed on the NYSE or Nasdaq.
To do so,
it plans to ramp up spending on advertising and other branding initiatives.
This
remains a long-shot proposition.
All these risks were detailed in
Archipelago's IPO documents, so the stock's
enthusiasts are betting that the
firm's well-regarded management can continue
its knack for staying a step
ahead of the industry's fast-evolving trends.
One opportunity is for Archipelago
to use its new currency to consolidate the
industry, most notably by merging
with Instinet or just buying that company's
ECN business, says Repetto.
In contrast to the market's willingness to bank on Archipelago, suspicion
pervades Knight Trading.
A Nasdaq market-maker and asset manager, Knight
rode the online-trading
bonanza of the late-'Nineties by handling the torrents
of orders from
discount-brokerage clients. Management made arrogant public
statements about its
purported no-lose trading techniques. The firm eventually
agreed to fines to
settle allegations it mishandled customer orders.
Though top management is entirely new and Knight's practices have changed for
the better, the firm still carries a taint among investors. Its shares have
fallen below 9 from 16 early this year. Part of the story is the steady slide
in
retail trading volumes and low stock-market volatility, which may or may
not
revive any time soon.
Yet Knight's shares are now being viewed by
some investors as a low-risk value
play and a potential beneficiary of a buyout
some time down the road.
Knight handles 15% of all Nasdaq-listed share
volume. Much of that is still
from discount brokers. But the firm is also
expanding its share of
institutional-order flow, currently about 2.5%.
The other part of the business is the asset manager Deephaven, a hedge-fund
group that manages $3.2 billion in relatively steady, arbitrage-type strategies,
and has a 15-year record of strong risk-adjusted returns.
Knight -- with
a $1 billion market value and $230 million in reported cash on
hand as of
June 30 -- has a $200 million investment in Deephaven and earns 50%
of its
income. The firm sold its options-trading business to Citigroup last
month
for $225 million in cash. Knight retains its 7% stake in the International
Securities Exchange options market, which is about to go public.
The bull
case goes on like this: Knight is now extremely conservative, taking
minimal
trading risks. The market-making business earned a small profit in one
of
the slowest quarters in years, this year's second, and has leverage to busier
or more jumpy markets. Net of its cash, the stock trades at less than 10-times
forecast 2005 earnings. And the company announced recently it will buy back $160
million in stock, or some 15% of its shares.
One important risk is that Knight
may decide to use its armored balance sheet
to make a major acquisition. Another
is that the pace of trading trails off even
more amid broad investor apathy.
Yet recent converts to the stock think it preferable, and more likely, that
at
some point Knight's new, shareholder-focused management may elect to sell
for a
double-digit share price to a player who thinks it smart to get much
bigger in a
hurry.
---
|
|
