As others have pointed out, the market hates uncertainty and so it responds
by applying an extraordinarily high discount rate. As value investors, we get
paid if we are able to identify a mismatch between the level of uncertainty and
the price that discounts a far more pessimistic scenario. My thesis is that a
number of steps taken by the company and the federal government in the last few
months have significantly reduced the likelihood of a level of financial
distress that market participants fear - i.e. those scenarios that would prevent
American Express from continuing as a going concern. Since the chance of
failure is greatly diminished and the likelihood of returning to a normalized
environment has now increased, American Express represents a good long-term
investment opportunity because it allows current investors to capture the high
discount rate as their expected IRR with the likelihood that it will be even
higher if the discount rate diminishes materially within the next two to three
years.
The steps that American Express has taken to counteract the three issues
above are the following:
- The company has announced very significant cost reductions. A few months
ago, American Express announced that it was cutting 10% of its workforce, as
well as significantly curtailing its investments. These cutbacks are expected
to net more than 800 million of savings in 2009 alone. These cost reductions
will not significantly impact revenues - so will increase profitability and/or
offset some of the revenue weakness.
- The company is also adjusting its pricing upward to offset some of the
pressures being created by credit losses as well as funding costs.
- A few days ago, American Express announced that it will be receiving $3.39
billion in the form of a preferred investment by the treasury. This preferred
stock will have a 5% yield to the treasury and 15% warrant coverage, which
represents the upside on approximately 28.25 million shares (strike price
presumably at current prices). This represents only 2.4% of the 1.16 billion
shares outstanding. This funding was received on very attractive terms, and
significantly increases AXP's ability to absorb significant additional credit
losses
- American Express became a bank holding company, specifically so that it
could accept funds from the treasury. In addition, American Express's status as
a bank holding company may give it flexibility to solve or mitigate some of its
funding issues through additional deposit gathering, and/or acquisitions of
banks with deposits. While this would not be an ideal outcome as a bank is not
nearly as good a business model as American Express already has, the flexibility
to solve a potential funding issue with greater number of tools is a positive as
it also materially reduces the risk of failure.
Regarding the three issues cited above, it is helpful to have a sense for
the scope of each concern:
Issue#1: magnitude of potential credit losses
Before driling into the specifics of the credit losses, it's important to
keep them in context, because American Express has a spend centric and not just
a lend centric business model - as pointed out above. This means that the
majority of their revenue is of a very profitable, fee based, growing, recurring
annuity type income stream.
Keeping aside the credit part of their business for the moment, AXP
generated $11.5 billion in revenue in the first 9 months of 2008 from this
Discount Revenue alone. I.e. the company's roughtly 2.56% cut of all the money
spent on their cards during that time period. During the 1st 9 months of 2008,
AXP cardmembers numbering approx 52.3 million in the US and 37.8 million outside
the US spent a total of $522.8 billion on their cards (This spending was up from
$469 billion last year due to both higher average spending per card member and
growth in cards-in-force despite the recession). In addition to the discount
revenue, the company also generated an additional $1.7 billion in revenue from
the annual fees that members pay to have an American Express Card (on average).
The beauty of this part of their model is that it is fee for service and it is a
toll bridge type revenue stream based on the growth of electronic payments and
global growth. This business requires very little capital, throws off lots of
free cash flow and is far larger than their lending business. As we will
discuss in Issue #3, this portion of their business is certainly sensitive to
the level of spending and in a consumer recession, growth will slow. It would
be unlikely to go negative for perhaps more than 1 or 2 quarters (and it has not
done so yet). In the other hand, nominal inflation will actually be fine as it
is likely that costs will not increase as fast as revenues.
Now compared to the total charged on the cards, the company's
recievables/loans to cardmembers outstanding are far smaller. As of Sept 30,
2008, total Cardmember Loans were $43.1 billion (this excludes approximately
$28.9 billion of additional cardmember loans in a Master Securitization Trust
which has been securitized, but includes AXP's retained undivided sellers
interest of approximately $13.5 billion in this trust) and card member
recievables (not loans) were approximately $40.6 billion. The latter figure
represents the delay between when the money is spent (paid to the merchant) and
when the customer pays American Express (i.e. it is reverse float). The former
are actual loans against which the provisions are $2.64 billion at the end of Q3
2008 - and this is the figure (loan provisions) that could and probably will go
much higher. One indication for the quality of the business is its ability to
absorb a higher level of loan losses. In Q3 2008, AXP provisioned $958 million
for loan loses (as compared to $579 million in Q3 2007). The Q4 provision will
probably be higher still. However, in Q3 2008, after absorbing higher loan loss
"reserves", the company still generated net income for $815 million.
Depending on the level of provisions in Q4 2008 and Q1 2009 which could
very will deterioriate from Q3 2008 levels, net income could go negative. The
company did expand its loan portfolio more aggressively that it should have in
2006 and 2007 and loan losses on part of its loans will be much higher than
expected. Even if they turn out to be 10% of loans outstanding (which would be
surprising), AXP has significant ability to absorb a temporary hit to
earnings.
Issue #2: Funding the balance sheet
AXP funds the aforementioned recievables and loans with short-term and long
term debt as well as customer deposits. As of Q3 2008, ST debt was $13.9
billion (down from $17.7 billion in Q3 2007), long-term debt was $57.7 billion
(up from $55.2 billion) and customer deposits were $11.8 billion (down from
$15.4 billion). This is a substantial balance sheet and parts of it frequently
have to be refinanced. Access to short and long-term funding in the commercial
paper, interbank, medium term notes and long-term notes were all impaired this
year. Recent actions by the Fed and the Treasury have significantly eased
liquidity. AXP has the ability to go to the Fed and borrow (probably the entire
amount) if it needed to. However, it appears likely that with Washington
focused on easing and getting lending going again, AXP will be the beneficiary
of policies designed to insure that refinancing risk is mitigated for companies
that have relatively viable businesses and assets (but that might run into
refinancing risk). Recently, the FDIC backstopped a debt issue for AMEX (by
providing a guarantee) and subsequently, AXP issued the aforementioned Preferred
Stock to the Treasury as well. Given current government policy, it seems highly
unlikely that AXP would be allowed to fail simply because markets are not
willing to finance financial companies for the time being. In the intermediate
term, of course, the cost of funding may go up, but AXP can pass along much of
this cost to its customers over time.
Issue #3: Earnings Hit
So the impact of slower spending and loan provisions and higher cost of
borrowing in a credit constrained environment are simply that margins compress
in the near term. Dec 2009 estimates decreased from $2.92 90 days ago to $1.92
today. However, AXP is not standing still. The company is simultaneously
increasing fees and interest charges, significantly cutting costs, and also
tightening credit standards. Meanwhile, spending and electronic spending in
particular will continue its upward March over time so the company is likely to
recover and go above peak earnings again. Earning may be lower in 2009 and
perhaps even in 2010. However, they should recover thereafter. In 18 to 24
months (at the outside), investors should be looking past current issues around
provisioning, borrowing, and earnings compression. Applying a 15x multiple of a
mimum of $3.00 in earnings in 2011 give a stock price north of $45.00 in 18 to
24 months for an IRR north of 30% per year from current levels.