Brief Japan: Shimadzu (7701 JP): 3Q Results Suggest a Trading Range and more

In this briefing:

1. Shimadzu (7701 JP): 3Q Results Suggest a Trading Range
2. Semiconductor Sales Dive A Record 7% MoM In December. 2019 Will Be A Low-To-No Growth Year.
3. Olympus: 3QFY03/19 Profits Decline on the Back of Litigation Related Costs
4. Softbank Buyback More Than It Appears To Be

1. Shimadzu (7701 JP): 3Q Results Suggest a Trading Range

Shimadzu’s 3Q results were good enough to reassure long-term investors, but not good enough to be called a buy signal. Sales and operating profit were up 4.5% and 4.6% year-on-year, respectively, in the three months to December, an improvement over 2Q but well below the double-digit increases recorded in 1Q and last fiscal year.  Forex losses and other factors led to a 2.2% decline in net profit. 

Sales were up in Japan, Europe and Asia ex-Japan and ex-China, but down in America,  China and Other Regions. Sales of core Analytical & Measuring Instruments were up 2.4%, operating profit on those sales was up 4.1% and the operating margin rose to +15.4% from +15.1% the previous year.

Sales of Industrial Machinery were down 5.7%, but operating profit on those sales was up 2.7% and the division generated a +9.7% operating margin vs. +9.0% the previous year. Sales of turbo-molecular pumps, primarily to semiconductor equipment makers, were down 14.3%.

Medical System sales were up 10.6% and the division generated a +1.5% operating margin vs. + 0.1% the previous year. Aircraft Equipment sales were up 12.1% but the division made a -0.5% operating loss vs. +1.2% profit the previous year. 

At ¥2,659 (Friday, February 8 closing price), the shares are selling at 24x our EPS estimate for FY Mar-19 and 12x EV/EBITDA. The five-year historical P/E range is 13x – 30x, the EV/EBITDA range is 6x – 16x. Over the next several quarters, we expect continued weakness in Industrial Machinery to offset single-digit growth in Instruments, keeping overall growth low. 

2. Semiconductor Sales Dive A Record 7% MoM In December. 2019 Will Be A Low-To-No Growth Year.

Global Semiconductor Sales for December 2018 amounted to $38.2 billion, down a record 7.0% MoM, according to the latest data published by the Semiconductor Industry Association (SIA). The December data reflects a sharp acceleration of a downward trend which began in November and comes as little surprise following an earnings season characterised by profit warnings led by industry giants such as Apple, Samsung and Nvidia. 

The December decline amounted to ~$3 billion in absolute terms, far less than the roughly $15 billion that failed to materialise in fourth quarter sector revenues and implying that substantial amounts of inventory still remain to be consumed from within the supply chain. 

As such we anticipate monthly semiconductor sales continuing to decline through April-May timeframe before stabilizing and returning to growth thereafter. We now anticipate growth to moderate significantly from the 13.7% experienced in 2018 to just 1% in 2019. 

3. Olympus: 3QFY03/19 Profits Decline on the Back of Litigation Related Costs

Olympus Corporation (7733 JP) reported its 3QFY03/19 results on Friday (08^th February) after markets closed. The third quarter revenue dropped 1.7% YoY while operating profit declined by a significant 21.5% YoY, which was 12% below consensus estimates. The operating profit margin for the quarter was 8.8% compared to 11.1% for the same period last year.

The cumulative nine-month results were not impressive either. Although revenue saw a marginal improvement of 1.6% YoY, operating profit declined by 66%, resulting in a 700 basis point decline in operating margin, which fell to just 3.5%. Revenue and operating profit missed consensus estimates by 0.4% and 10.4%, respectively.

The company shares are currently trading at JPY4,645 per share which we believe is overvalued based on our EV/EBIT valuation. The premium is not justified given the governance related issues and the scandals currently faced by the company. Further, Olympus’ financial performance has been disappointing recently, and the company’s largest segment is growing only at single-digits and the Imaging business continues to drag on company revenue and margins. The share price gained nearly 38% since the beginning of the year following the company’s announcement to transform its business and improve governance. In our view the potential for a transformation in governance and business practices is already fully-discounted in the share price.

4. Softbank Buyback More Than It Appears To Be

Softbank Group (9984 JP) last week announced its Q3 results. The stock popped 15+% quickly that day and stayed up all day long, closing at +17.7%. The next day was up small. Over the two days volume was 74.1 million shares. 

I expect the shares were up for two reasons.

1. People figured out Son-san could sell as well as buy. And the sale of NVIDIA Corp (NVDA US) shares was done very well. 
2. Softbank announced a buyback of ¥600 billion – its largest buyback ever. 

The first was surprisingly well-executed. The ownership and transfers of assets from Softbank to the Softbank Vision Fund are sometimes tough to follow, and this should give non-Softbank SVF investors some pause, but Softbank’s ability to get leverage on assets is good, and the collar transaction was – in the eyes of this former derivatives strategist – very well done.

The US$15bn+ gain in market cap over the next two days was probably 10 times the net income impact of the savings on the NVIDIA trade, which means investors are paying 10x earnings for the same thing to incrementally happen every year vs what they thought was going to happen before Thursday. Financial trading businesses have generally traded over time in the high single digit PERs because of the variability of results, so the jump was a little more than it should have been for that, especially if you think some years the “right” jump because of better-than-market execution will have less impact than $2.9bn.

So the rest was either due to other business going well, or the share buyback. At ¥600 billion and at Friday’s closing price, it is about 7 days worth of volume using the 3-month volume average prior to the earnings release and 8.6 days of volume using a one-year average. That means they could buy 10% of ADV every day for 70-86 trading days and complete the buyback, or it means they could buy 3.4% of the volume every trading day. 

That doesn’t seem like a lot. But it is.