4 Airline urban myths.........................Salomon Smith Barney ~ June 12, 2001
Airlines
We Analyze Four Urban Legends (For Airline Stocks)
June 12, 2001 SUMMARY
* Taking a brief break from our earnings cuts, we analyze
four popular beliefs on making money in the airline sector.
Namely that airline stocks outperform: 1) six months prior to
an economic recovery; 2) six months prior to a recovery in
airline earnings; 3) during periods of falling interest
rates; and 4) that airline stocks should be purchased in the
winter and sold in the summer. *We find scant empirical
evidence to support numbers 1, 2 or 4 above. *The results for
item 3 were more interesting: Airline stocks appear to
outperform nicely (Big 3 cyclical majors +14% vs. S&P400 on
average) between 3 and 6 months immediately following a peak
in interest rates. However, impeccable timing is required,
with the airlines marking time in the three month period
immediately after the interest rate peak as well as between 6
and 9 months after the peak. *Meanwhile, sector valuation
looks unexciting (market well conditioned not to sell on bad
news) and we remain about 25% below consensus 2002 earnings.
FOUR AIRLINE STOCK URBAN LEGENDS
While covering the airline sector, we have encountered a number of investor
rules-of-thumb which have swelled to near mythic proportions. In this report
we investigate four of these themes to see if the legends are supported by
empirical data: 1) airline stocks outperform the market six months prior to
an economic recovery; 2) airline stocks outperform six months prior to a
recovery in airline earnings; 3) airline stocks outperform during periods of
falling interest rates; and 4) airline stocks should be purchased in the
winter (anticipating strong summer earnings) and sold in the summer (prior to
the seasonally weak winter period)
.
LEGEND 1: AIRLINE STOCKS OUTPERFORM SIX MONTHS PRIOR TO AN ECONOMIC RECOVERY
To gauge the merits of this assertion, we isolated six instances over the
past 15 years where year-over-year quarterly U.S. GDP growth increased by
over 1% point vs. the prior sequential quarter (e.g. 1Q year-over-year GDP up
1% while 2Q is up 2%+). We then looked at the stock performance for the Big 3
U.S. cyclical airlines (arithmetic average performance for AMR, UAL, and DAL)
within each of the three quarters prior to the quarterly uptick in economic
growth. As shown in Figure 1, airline stocks did indeed outperform the market
(defined here as the S&P400) during the quarter six months prior to the
economic uptick, but by an unexciting margin of 5.8% to 4.8%. The
outperformance was slightly stronger looking to the quarter nine months prior
(+9.0% vs. +7.4%) while the airlines actually underperformed during the
quarter immediately prior to the economic uptick (+3.8% vs. 4.7%). The
results ranged widely in each of the periods, further limiting the
effectiveness of this bit of stock-picking lore (from -13% to +21% prior 3
months, -7.0% to 29.1% prior 6 months, and 12.0% to 31.9% prior 9 months).
(N.B. As with all four of the airline stock urban legends which we analyze,
if you slice and dice long enough--changing the study period, classification
for significant uptick in economic activity, relevant sample of airline
stocks--eventually you may find the expected correlation.
In each case,
however, we took what for us is the most logical cut we could isolate given
data limitations.)
Figure 1. Average 3 Month Returns for the Big 3 and S&P400 When GDP Increased
By More than 1% Point (1985-2000)
If GDP Up If GDP Up If GDP Up If GDP Up If GDP Up If GDP Up
1%+, 1%+, 1%+, 1%+, Return 1%+, Return 1%+, Return
Return 1Q Return 1Q Return 2Q 2Q Before 3Q Before 3Q Before
Before Before Before (S&P400) (Big 3) (S&P400)
(Big 3) (S&P400) (Big 3)
Avg 3.8% 4.7% 5.8% 4.8% 9.0% 7.4%
Return:
Under/Over -0.9% 1.0% 1.6%
Perf:
Note: Nominal GDP returns are compared with nominal returns in stock prices.
Source: Factset and Salomon Smith Barney.
LEGEND 2: AIRLINE STOCKS OUTPERFORM SIX MONTHS PRIOR TO A RECOVERY IN AIRLINE
EARNINGS
This theme is similar to the previous one, except instead of anticipating
improvements in GDP, we analyze whether airline stocks anticipate a
significant uptick in the airlines' own earnings. For this analysis, we
isolated 6 quarters from 1990-2000 where year-over-year earnings for the Big
3 improved by at least 25% when they had not shown similar or better
improvement the prior sequential quarter (thereby isolating the inflection
point when earnings substantially improve). Again, we looked at the stock
performance for the Big 3, one, two and three quarters prior to the uptick.
As shown in figure 2, the Big 3 outperformed the general market during the
three months prior to the significant earnings improvement (Big 3 up 8.3% vs.
S&P400 up 6.0%), but lagged in the quarter 6 months prior and in the quarter
9 months prior by 0.4% and 1.3% respectively. Once again, returns varied
widely within each of the periods prior to the 6 inflection points (0.1% to
16.7% quarter immediately prior to improvement, -12.0% to 31.9% during
quarter 6 months prior, and -11.2% to 19.1% during quarter 9 months prior),
suggesting weak statistical correlation.
Figure 2. Average 3 Month Returns for the Big 3 vs. the S&P400 When Quarterly
Operating Income Increased By More than 25% Points (1990-2000)
If O/I Up 25%+ If O/I Up If O/I Up If O/I Up If O/I Up If O/I Up
1Q Before 25%+ 1Q 25%+ 2Q 25%+ 2Q 25%+ 3Q 25%+ 3Q
(Big 3) Before Before Before Before Before
(S&P400) (Big 3) (S&P400) (Big 3) (S&P400)
Avg. Return: 8.3% 6.0% 5.2% 5.5% 2.5% 3.8%
Under/Over 2.3% -0.4% -1.3%
Perf:
Source: Factset and Salomon Smith Barney
LEGEND 3: AIRLINE STOCKS OUTPERFORM IN PERIODS OF FALLING INTEREST RATES
As with Johnny Appleseed (OK not exactly an Urban Legend), there actually
appears to be some truth underlying this particular story. Here we isolated
six discernable interest rate peaks (defined here by the daily effective
funds rate) over the past twenty years and compared the relative 3 month
performance of the Big 3 airline stocks three, six, and nine months following
each peak. As shown in figures 3 & 4, if you bought airline stocks three
months after interest rates peaked and sold three months later, you would
have outperformed the market on average by 13.9% (absolute performance of
+20.0%). Not bad. However, you'd best be nimble. The airline stocks basically
marked time during the three months immediately following the peak in
interest rates (outperforming by 0.4%) and then during the subsequent six to
nine month period following the peak (underperforming by 0.9%). Once again
results varied widely (from outperforming by as much as 43% and
underperforming by 10% in the three to six month period).
Figure 3. Summary: Peak Interest Rates and Big 3 vs. S&P 400 Stock
Performance (1980-2001)
Avg 3 Mo. Avg 3 Mo. Avg 6 Mo. Avg 6 Mo. Avg 6 Mo. Avg 6 Mo.
After Peak After Peak After Peak After Peak After Peak After Peak
(Big 3) (S&P400) (Big 3) (S&P400) (Big 3) (S&P400)
Avg. 1.7% 1.4% 20% 6.1% -3.2% -2.4%
Return:
Under/Over 0.4% 13.9% -0.9%
Perf:
Note: Returns assume a 3-month holding period 3, 6, and 9 months after peak
interest rates and are not cumulative
Source: Factset, Stockval and Salomon Smith Barney
Figure 4. Detail: Peak Interest Rates vs Over/Under Performance of Big 3
Interest Rate Relative (Over/Under) Performance
Peaks 3 months 6 months 9 months
Mar-80 -17.3% 17.9% -11.4%
Jun-81 -18.3% -9.7% 27.5%
Aug-84 18.3% 13.7% 7.4%
Feb-89 4.9% 42.9% -25.7%
Mar-95 14.2% -0.7% -0.1%
Jun-00 0.5% 19.5% -2.8%
Average 0.4% 13.9% -0.9%
Note: Returns assume a 3-month holding period 3, 6, and 9 months after peak
interest rates and are not cumulative
Source: Factset, Stockval and Salomon Smith Barney
LEGEND 4: BUY AIRLINE STOCKS IN THE WINTER AND SELL THEM IN THE SUMMER
The reasoning here is investors should purchase airline stocks during the
seasonally weak winter period in anticipation of strong summer earnings and
conversely sell the stocks when earnings are strong in the summer (we suspect
this approach is popular with January effect devotees). Turning to figure 5,
we looked at the relative monthly performance for the Big 3 (AMR, UAL, DAL)
airline stocks since January 1990. While more talented entrail readers may be
able to discern clear patterns from the data, we cannot. Airline stocks
outperformed three of the six seasonal trough months (October-March) as well
as three of the six seasonal peak months (April-September). Not surprisingly,
the results ranged widely for each month (e.g. -7.8% to 18.2% for the
statistically strongest month March).
Figure 5. Number of times the Big 3 Out/Underperformed the S&P400 by month
(1990 - present)
Over/Under Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Performed
# Times 3 7 10 8 2 6 7 1 4 8 4 6
Outperforme
d
# Times 9 5 2 4 9 5 4 10 7 3 7 5
Underperfor
med
Average -2.8% 0.5% 5.5% 1.3% -3.8% 0.1% 0.6% -5.6% -1.6% 3.6% -0.5% -0.6%
Over/Underp
erformance
Source: StockVal and Salomon Smith Barney
SECTOR OUTLOOK: STUCK IN THE MUD
Frankly, we're having a hard time getting traction on the sector. The current
revenue outlook is dreadful due to the most severe case of business buydown
on record (business travelers booking ahead, putting up with Saturday night
stays, scrounging on the Internet etc. to take advantage of lower fares, no
doubt incentivized by the recent record gap between business and leisure
fares). Yet as analysts repeatedly cut numbers the market appears to have
enthusiastically embraced the mantra not to sell the stocks when the outlook
is bleakest. Consequently, the sector has yet to become cheap enough for us
to become enthusiastic (sector currently trading at .24 2001 Price/Sales vs.
trough average .20 and lowest trough 0.14 October 1994). Meanwhile, 2002
consensus numbers appear aggressive, implying the greatest increase in pre-
tax profit in the past decade. We're about 25% below 2002 consensus; even so
we're showing a 3.1% point increase in operating margin, recouping most of
the 4.2% point anticipated decline for this year.
We would look forward to taking a more aggressive stance on the sector should
the shares retrace another 15% to get to .20x 2001 Price/Sales, equaling the
average trough multiple. Failing that, we'd at least like to see consensus
numbers fall to a more reasonable range (defined of course by our own levels
25% below consensus 2002). (While we understand that the airline sector may
not be unique in sporting overly optimistic 2002 expectations, airlines tend
to lag on both ends of the economic cycle--thanks to sticky corporate travel
policies. Hence airlines--and airline stocks--may continue to languish as
other cyclicals begin to perk up.) Within this anemic industry outlook, we
continue to favor NWAC (labor unrest behind and 2002 benefits from new
infrastructure in Detroit), Continental (good valuation and strong hubs), and
(for deep value hunters only) UAL.
Figure 6. Looking ahead at Projected YOY Growth in Industry Operating Income
for Airline Industry
Quarter SSMB Proj GDP YOY Estimated Change in
% Change by Qrtr YOY Op Income*
2Q01 1.9% -83.4%
3Q01 2.5% -22.1%
4Q01 3.5% 172.4%
1Q02 4.3% n/a
2Q02 4.7% n/a
3Q02 3.7% n/a
4Q02 3.8% n/a
*4Q01 Growth coming off of small base. Source: Salomon Smith Barney. |
