Description

In January 2011, Meg Whitman was hired as CEO to turn around HPQ. As part of the turn-around effort, HPQ will spin-off its Enterprise business (HPE) on November 1^st, and leave behind its shrinking PC/Printer business (HPI). On September 15^th, HPQ will host an analyst day to discuss the spin-off.  We think the stock of the Enterprise business will trade to a higher multiple than the combined company, and that the stock of the declining PC/Printer business should be supported by a high free cash flow/dividend yield.  On a combined basis, we see +15% upside before year end, and relatively limited near term downside because (1) HPQ will conclude the spin-off before it makes its next quarterly report in late November and (2) the stock has fallen -27% from the time the spin-off was announced in October 2014 and now trades on a relatively low 4.5x TEV/2015E EBITDA.  Market participants may adopt the habit of valuing HPE’s finance business separately, and that would allow for greater upside to our base case.

Scenarios:

Bear case (no spin, -5% to -10%): While there is always the unknown in any investment, a cancellation of the spin appears unlikely. From a regulatory perspective, the SEC and IRS have no problem with giant operating companies such as HPQ splitting into two pieces in order to make each piece more manageable and easier to understand. The Form 10 has already been through two iterations with the SEC. From the board’s perspective, the split is a critical part of Meg Whitman’s turn-around plan. Operationally, the companies have already completed the physical separation. See below:

Meg Whitman (CEO): “On August 1st, we successfully split the operations and IT systems for the company. This was an incredibly complex process, and the team executed very well. This separation required working directly with more than 3,500 customers and partners to prepare for the cutover. We successfully separated nearly 750 systems affecting 95% of our business with no issues. After shutting down for just three days to transition, critical operational systems across our business segments are now live globally. Customer orders are flowing through manufacturing, and shipments are in transit across our entire supply chain network. This was a huge accomplishment.” (Q3 2015 earnings call 8/20/15)

Base Case (spin and re-rating, +15% to +30%): The Spin takes place November 1st, and HPE trades up because it’s no longer inhibited by a declining PC/Printer activity.  HPI shares are supported by a meaningful dividend yield and management’s willingness to return free cash flow to shareholders through buybacks.

Bull case (spin and M&A, 30%+): The spin takes place Nov 1st, HPI.  Receives an LBO bid in December, and then HPE engages in a strategic M&A of some form (or reports organic growth).  In September of 2014, the Wall Street Journal reported that HPQ and EMC had been in merger discussions off and on during the prior twelve months.  HPI would be an attractive LBO target for a large private equity firm as a declining cash generating business with limited leverage and an established franchise.

The Turn-around of Enterprise:

The table below shows HPE’s change in revenue, by segment, and excludes the effects of foreign currency fluctuations. Three of the eight reporting segments account for roughly 75% of sales: Enterprise Services (36.4%), Industry Standard Servers (24.4%) and Technology Services (13.7%). Compared to the year over year growth in each of the past two quarters, Enterprise Services (-3%), got significantly less-worse, Industry Standard Services (+15%) grew nicely and Technology Services (-3%) got a tad worse. HPQ acquired Aruba Networks in May of 2015 and that accounts for the 28% increase in Networking segment revenues in 3Q15A.  Without the acquisition, Networking revenue growth would still have been in the mid-single digits ex-currency.

Since the 1930s HPQ has managed technological change and the related shifts in its mix of sales.  At a high level, the broad adoption of cloud computing means that HPE customers are removing a layer of local server equipment, and instead connecting workstations and other devices to networking gear that in turn communicates directly with remote server facilities (i.e. the cloud).  HPE revenues may have reached a long awaited inflection point, but based on three consecutive quarters of -2% currency adjusted revenue declines, one has to take the CEO at her word:

Meg Whitman (CEO): “We're now rounding the corner where the growing businesses are bigger than the declining businesses, which is actually going to lead to growth. And boy, that has taken a while to get here, but we're here. And I think that sets Hewlett-Packard Enterprise up pretty well as we think about growth going forward.” (Q3 2015 earnings call 8/20/15)

If HPE showed say a +2% improvement in organic growth instead of a -2% decline this past quarter, but there was no spin-off in the works, it’s unclear that anyone would have cared. With the spin-off around the corner, HPE now has the potential to command a higher multiple for successful execution of the turnaround whereas before it did not.

In May 2015, HPE undertook a second transaction (aside from the acquisition of Aruba Networks) and agreed to sell 51% of H3C Technologies to Tsinghua Holdings, the asset management arm of Tsinghua University in China. H3C Technologies is HP’s China based server, storage and technology services business. The $2.3bn sale may close after the spin-off but before year end. We leave it out of our valuation because we’re not sure how much EBITDA will be going with the sale of the 51% interest.

Sum of the Parts:

After the investor day, we should have a better idea of the EBTIDA and free cash flow generated by each company. The following attempt should be roughly right but precisely wrong, so to speak.  We directly calculated HPE EBITDA from the latest Form 10 for 2014, and fiscal 1H15 ended April. Then we subtracted the overall company consensus EBITDA from the HPE EBITDA to arrive at a crude estimate of HPI EBITDA. From here we looked at the percentage allocation of EBITDA to each segment, and then applied that percentage allocation to the current 2015 consensus to arrive at EBTIDA estimates for each segment. Because roughly $10.6bn of the $16bn of pro-forma debt at HPE should relate to its finance business, for now we added back to EBITDA the net interest expense associated with the finance co.

HPQ disclosed in its Form 10 that the spin-off will create $400-$450mm of dis-synergies related to corporate functions such as finance, legal, IT, real estate and HR.  The dis-synergies will be evenly divided between HPI and HPE according to the company. In the table below we reduced each EBITDA estimate by $450mm/2= $225mm. We’ve seen a number of third party 2015E estimates for HPI and HPE EBITDA, respectively, some that are higher and some that are lower. Again, this is imperfect.

Below please find a comp table that depicts the peer company averages for each HPQ segment.

The average stock in the HPE peer group has traded down -0.4% since the announcement of the spin-off in October 2014, and the average stock in the HPI peer group has declined -24.7%. Over the same period, HPQ has fallen -26.8%. It would appear that HPQ has traded in line with the HPI peer group, independent of the fact that the majority of HPQ’s EBITDA, and half its pre-tax operating profit, comes from HPE.

HPQ trades on 4.5x Adjusted 2015PF EBTIDA.  If we assume HPE trades at a slightly higher multiple than the combined company and that HPI should trade at a lower multiple than the combined company, then we get the following decomposition based on the capitalization disclosed in the Form 10. HPE makes up about $18.60/share at 5.0x and HPI contributes $8.50/share at 3.6x.

The 5.0x multiple on HPE is well below the 7.6x peer average and the HPI 3.6x multiple is also below its respective 5.1x peer average. Some argue that HPQ should trade at a discount because HPI is in decline and HPE has lower margins than its peers. HPI is struggling with a declining ink and toner business and the consumer substitution of hand held devices/tablets in place of desktop computers.  HPI also faces increased competition in its printing sub-segment from Japanese rivals able to compete on price thanks the depreciation in the Yen. We’ll be looking for an update on these trends at the investor day.

The 2013 DELL LBO was done at 4.9x LTM EBITDA. Dell’s consumer business accounted for 20% of sales at the time of the LBO but was unprofitable on an operating basis. At HPQ, HPI’s segments are not solely consumer facing and contributed roughly half of sales and (as mentioned above) half of pre-tax profits in FY2014A.  In particular, printing supplies account for about ¾ of HPI’s operating profits. So, the Dell LBO multiple may only serve as a remote proxy for the LBO multiple on a prospective HPI take out.

HPQ expects the combined dividend payment to remain around $1.1bn post-spin, with HPI set to pay out somewhat more than HPE. So, if HPI paid out say $765mm and had $15.3bn of equity value, as implied in the table above, that would translate to a 5% dividend yield. If we leave HPI on 3.6x in the belief the market has priced in support from a prospective dividend and move HPE up one turn to 6x to reflect a potential turn-around, the analysis generates $4.39/share of upside ($31.49/$27.10-1=16%).

              The HPQ captive finance business generated $124mm of pre-tax net interest income in FY2014 and as of F3Q15A is annualizing at a $121mm pace for FY2015E. Note that HPQ ends its fiscal year in October and that the HPQ finance business will go with HPE in the spin-off. If we remove the financing income from our HPE EBITDA estimate, remove the cash and debt used by the financing business, and add to the market cap the book value of the equity of the finance business, then we would see a different picture (below).

              Notice that we subtracted from HPE $600mm of finance co related cash and $10bn of finance co related debt, and reduced HPE EBITDA by $121mm.  We valued the finance co at 1x book value, equivalent to $0.83/share.  ROE looks to be in the 7-8% range and comparable stand-alone companies trade at about 1x book.

              By mixing valuation methods (P/B v and TEV/EBITDA) we moved $10bn of net debt out of the analysis and made the EBITDA multiples on the operating business appear lower.  We assumed HPI would pay a dividend of $765mm (from the prior example) and then backed into a share price that implied a dividend yield of somewhat less than 10% (9.1% in this case). That yield corresponded to 2.0x for HPI EBTIDA, and further implied an HPE multiple of 4.5x.

The tables below illustrate the HPQ share price at varying prices and multiples for HPI and HPE, respectively. The first set of tables below, wherein we back out the captive finance business, will look about twice as favorable as the second set of tables. The second set of table are driven with a more basic initial approach. Although the first set of tables will look better, it’s worth noting that HPQ is trying to add value by separating HPI and HPE, not by separating the finance activity from HPE. The finance business has strategic value because it allows HPE to compete on the basis of the timing and amount of payment, not just the amount of payment (i.e. price). Now that HPE will be independent and easier to understand, the major uncertainty is whether market participants will adopt the habit of separately valuing the finance co. None of the companies on the comparable table, aside from HPQ, break out a separate finance division, so perhaps it’s reasonable to think people will notice this difference and adjust accordingly.

The first two tables below back out the effect of the finance co on TEV and EBTIDA, respectively, and we add a constant $0.83/share benefit to each HPE share price. The left axis shows the HPI multiple and share price and the top axis shows the HPE multiple and share price.

The next two tables depict the simplified approach that leaves in the captive finance company:

              A word of caution, the output of the table is only as good as the input and the input should get substantially more accurate after the investor day next week.

Pensions and NOLs:

HPE as of 2014A had $5.5bn of NOLs ($260mm federal, $46m state and $5.2bn foreign). HPQ net of HPI, as of 2014A should have $29.3bn of NOLs ($598mm federal, $4.1bn state and $24.5bn foreign).  When we take the present value of the NOLs we get 15-20c on the dollar, or approximately $5.0bn at HPI and $1.2bn at HPE.  On a per share basis that means HPE would have $0.67/share of NOLs and HPI would have $2.78/share.  We could have treated the present value of the NOLs like cash and reduced the TEV of HPE and HPI by corresponding amounts, but that would just have served to make something very cheap look ridiculously cheap. We get it, it’s cheap. Next we could adjust for HPE’s underfunded pension ($2bn) - but that would more than offset the benefit of their NOL in economic terms, and of course the degree of underfunding could disappear if the return on investment improves.

The HPI pension plan (calculated as HPQ 2014 10K – HPE Form 10) was $3.8bn underfunded.  If we netted the underfunded status of the HPI plan against the NOL that takes the combined economic benefit of the NOL and the pension down to ($5.0bn - $3.8bn) = $1.2bn or $0.67/share. Here again, we’re subtracting a 2014A 10K number from a 2015PF Form 10 number to get a value.

Conclusion:

              On September 15^th, HPQ will host an analyst day to discuss the spin-off of HPE. On November 1^st, the spin-off will take place.  HPQ/HPI will report fiscal fourth quarter earnings in late November and at some point following the spin-off will publish stand-alone HPI pro-forma financials.  In the absence of clean disclosure around EBITDA and free cash flow at each business we’ve tried to err on the side of caution and to be clear about our assumptions and their related limitations. Since the spin-off was announced in October 2014, HPQ has traded down -27% in line with the HPI peer group and independent of HPE’s material profit contribution.

Our base case (+15% to +30%) assumes that HPI is supported by a 5% dividend yield and HPE gains 1x of multiple on the prospect of a turnaround. If we back out the captive finance company the risk/reward improves for the base case and the upside is closer to 30%.  The bear case (-5% to -10%) seems all but finished given that the two companies have physically separated, and will not report another quarter until after the separation. The bull case (30%+), involving the LBO of HPI and some deal activity in HPE (or organic growth), has some basis given the Dell LBO and credible reports of merger talks last year with EMC.  However, we’d rather wait to see free cash flow figures for each business to ground whatever premium EBTIDA multiple we might apply to quantify the bull case.

Catalyst

On September 15^th, Hewlett Packard will host an investor day in New York to explain the spin-off.

On November 1^st, the company will spin-off HPE, the Hewlett Packard Enterprise Company.