Friday, October 28, 2011

United Parcel Service (UPS)

The largest delivery company in the world. It is still quite dependent on the US: 60% of revenue and 57% of profit came from domestic operations in 2010.

Their segmental results are broken down into:
- US Domestic (ground delivery of letters or packages, both sending and receiving in the US)
- International Package
- Supply chain and Freight (Truckload (TL) and less-than-truckload (LTL) deliveries) - in the US and surrounding countries)

Competitive Advantage

DHL wrote off 3.9bn when it exited the US domestic market in 2009. Fedex and UPS are the only 2 large players left, and are compared below.

First, market share. UPS has larger (about 80% more) revenue for US deliveries. They both have similar revenues for International.
'International' here means the package was sent to *or* from the US.

Next, margins. This is difficult because both companies break down their segments differently. Fedex breaks down their margins into 'Express' vs 'Ground' (disregards if the packages are domestic or international):
We can see that:
  • Ground handling has a lot higher margins than express.
  • Express margins are more cyclical: were worse hit by the 2009 recession, due to falling revenue (Express fell 8% from 08 to 09, vs a 4% increase for Ground).
UPS breaks up their segments differently. Each of their 'Domestic' and 'Intl' segments is a combination of express and ground. Their 'Domestic' was an even mix of express/ground in 2010. Express has made up a larger and larger proportion of the mix since 2007.

We can't compare directly to Fedex due to the segment differences.

UPS has significantly higher margins overall. Morningstar says this is probably because of 1) their integrated network (same network used for both express and ground delivery, unlike Fedex, which uses separate networks), and 2) more mail being funneled through the network.


For the last 10 years, from 2001 onwards, UPS has generated free cashflows all except one year.

Balance Sheet

Long term debt on their 2010 balance sheet is minimal (about 10b), together with some others (e.g.: post-retierment benefits), totals about 18b. Thats about 5-6 times their 2010 income (3.4b).

Off the balance sheet: they have another 3bn operating leases and purchase commitments (included in income statement), plus 2.5bn capital lease (not in income statement). All spread out over the next 10 years.

Industry Outlook

Depends on trade (national and international).

In the short term, this depends on the economic cycle.

In the long term, international trade has been increasing since WWII. It would take drastic scenarios for trade to go down (e.g.: trade war, rise of additive manufacturing - a possibility explored by DHL).


Company with a large moat. Half US, half international. Highly cyclical. Buy in a recession.

Friday, October 21, 2011

Defined Pension Plans

Many US companies have underfunded pension plans, which are not shown on thier balance sheets. In a good case, its several years of earnings. In a bad case, the company may be bankrupt.

We are looking at 'Defined Benefits Pension Plans' where the company is liable for providing stipulated retirement benefits for employees at a future point in time. These are unlike the 'Defined Contribution Benefit Plans', where the employee/employer contribute a fixed amount (401K plans in the US, similar to CPF in Singapore). Liabilities for defined benefit pension plans may be quite big, are not on the balance sheet, and therefore ignored by the financial ratios (eg: PE, debt/Equity).

How are Pension Costs Calculated

Aim here is to understand the meaning behind the pension numbers in the financial statement, process and assumptions that go into the numbers we read. Here is how I imagine it they are calculated (I am not an accountant).

Step 1: Get the current 'fair value' of pension fund assets: This is quite straightforward and objective, as most pension plans consist of liquid assets (equities or bonds).

Step 2: Calculate the Projected Benefit Obligation (PBO): really a series of guesses upon guesses:
  • Estimate how much will be owed in future. There is a lot of guesswork here based on the employees' pay rises, retirement age, etc. Some terminology: changes to the final PBO as a result of these assumptions changing are labelled as 'Actuarial loss/gain' or 'Experience loss/gain'.
  • Reduce this to a number in the present by applying a 'discount rate'. This gives the PBO: an estimate of how much you need today to pay off your future obligations.
Step 3: Compare these numbers from steps 1 and 2: Is the plan under or over funded? 'Funded status' must be stated n the footnotes, but if underfunded, the liability does not have to be recognized on the balance sheet...hide it it the footnotes somewhere.

Step 4: Calculate how much the pension cost in the current year, then decide how much of that to expense in the income statement.

Calculate the current pension cost. This appears in the footnotes, something like 'Components of Pension Expense' or 'Net Periodic Benefit Cost'.... It consists of:
  • Service cost: Because all employees are now 1 year closer to retirement.
  • Interest cost: Because we now have 1 year less to payout our obligations. This depends on the 'Discount rate' in Step 2 (higher rate --> lower PBO --> higher interest expense every year).
  • Subtract the Expected return on assets. This can be increased/decreased.
  • The result is how much you expect the pension to cost you every year.
But....this amount does not have to be fully recognized in the income statement. It may be smoothed out (exceptional gains or losses can be deferred over five years).

In summary, its quite a complex model to reduce a set of future estimates into a few simple numbers, and in the end those numbers may not even be added to the income statement or balance sheet.

Summary of Things to Check

From Investopedia:

1) Locate the 'Funded status', if underfunded, is the liability recognized on the balance sheet?

2) Check the assumptions used:
  • Rate of salary increase used to calculate the graph in Step 2.
  • Discount Rate (step 2): Is it reasonably low? I'm not sure....what is a reasonable expected annual compound growth rate for equities or bond investments?
  • Expected Return On Assets in Step 3: Should not be significantly higher than the discount rate.
3) Is the pension cost/expense fully expensed in the income statement, or deferred?

4) See what proportion of the plan is funded by equities and bonds. Bonds are safer: as they are held to maturity, the only risk in holding them is that the underlying company goes bankrupt. Equities may rise or fall in value when the time comes to sell them.

A look at UPS

I'm using United Parcel Services (UPS) as an example, simply because its the next stock I'm looking at. Numbers are from its 2010 annual report. Page numbers refer to the page number in Adobe Acrobat (not the numbers printed at the bottom of the page).

Funding Status:
  • (p89) UPS' pension plan is worth 20m886m, with a BPO of 25,619m. As calculated, it is underfunded by 4.7bn. And all of this underfunding is recognized on the balance sheet (p90).
Assumptions used (mostly on p87 'Actuarial Assumptions'):
  • UPS has been using between 6.5-7% for the past 3 years. This may be acceptable: to me its not conservative, but not high either. p60 states that a 0.25% decrease in the discount rate results in a 850m increase in projected Pension Plan obligations.
  • Return on assets: UPS has used between 9%, and 8.75% in the past 3 years. Seems high, either for equities or bonds. p60 states that a 0.25% decrease in ROA will increase costs by 46m/year.
  • Rate of salary increase. UPS has used 4.5%, seems realistic.
There is an interesting note on p60:

"Our 2011 pension expense will be higher than our 2010 expense due primarily to two negative factors: the decline in discount rate used to determine expense from 6.58% for 2010 to 5.98% for 2011, (good: 6% discount rate is conservative) and the required amortization of unrecognized losses, the majority of which relate to 2008 asset losses, outside of the corridor we utilize for accounting purposes. These negative factors are partially offset by the additional discretionary contributions that we made in 2010 and 2011 that increased the expected return on assets used for expense calculation purposes." (Did they buy equities....? probably not good to buy high in 2010 and 2011 to make up for losses in 2008.)

Is the pension cost fully expensed?
  • The Net Periodic Cost is calculated at $903m (add columns on p87). This was fully expensed in 2011 (p69 cashflow statement - the amount is added to CFO i.e.: deducted from earnings).
  • I think p90 shows a buildup of pension costs from past years (total $5900m in 'Amounts Recognized in AOCI'). Not sure of the terminology - I think it means this just has to be recognized in future years, which is what they were talking about above (interesting note on p60).
Conclusion for UPS Pension Plan:
  • Can't find anything wrong.
  • Wait for the 2011 Annual Report, to see what the BPO looks like with a more realistic 5.98% discount rate, and how much of the of the $5900m accumulated costs (loss) they recognize.

Tuesday, October 4, 2011

Added more money

Added 260K. From sale of overseas apartment. Now have 510K.

If we get a recession, and I am able to buy stocks which don't go to zero, I am going to be rich.

Sunday, October 2, 2011

ECRI calls Recession

New recession call from ECRI. At a minimum, a shallow one, but if there are unexpected shocks, it would be deeper.

Continuing the chronological listing of ECRI's calls:
  • 31 Aug 11: Summary of Member Report Issued Aug 19, 2011
    Job Market to Remain Weak. Still cant tell if slower growth or recession. Has been persistent, not pronounced or pervasive. No upturn in sight yet.
  • 17 Sep 11: Radio interview: Skating on this ice.
    Fwd looking indicators show continued slowing. Risk of a new recession is "quite high", don't yet know. Should know by end Nov.
  • 30 Sep 11: US Recession (Bloomberg)
    Recession "now inescapable". Over a dozen 12 US leading indexes (different aspects in the US economy), all showing contagion. Vicious cycle to start. Will continue to be pronounced, pervasive and persistient. "If Europe cleans its act up and everything is good, there's still going to be a recession." If there is an unpredictable event (eg: Lehman imn 08) it will be worse. How long? Don't know yet. Right now, minimally, its a shallow recession. Recessions kill inflation.
Good. Soon I'll get a change to use the money I've been hoarding.