Investigations by DRAMeXchange, a division of TrendForce, show peak season demand growth and, in turn, production volume growth for the smartphone market in 3Q weakening in comparison to previous peak-seasons, for which production volume growths of 10% QoQ and above were the norm. The decline for total smartphone production this year is forecast to remain near 5%. Pull-ins have slowed in 1H, and we see suppliers yet to fully clear their large levels of inventories. Despite the Japan-South Korea incident over material exports, which resulted in rumors of price trend reversals circulating in the market, pressure to clear inventories still remained high. Adding the fact that though some DRAM suppliers have announced plans to reduce production capacity, the actual extent of those reductions were generally low,consisting mostly of capacity reductions for older processes or capacity reductions as a result of process switching. Until mainstream products become unprofitable, DRAM suppliers are unlikely to make large reductions to capacity. And though prices in spot markets have seen slight fluctuations, contract prices remain on a downwards trend overall.
TrendForce points out that mobile DRAM prices had to wait until the second half of July to stabilize amid a great many messages circulating in the market that make it difficult to provide quotes. As suppliers generally possess higher inventory levels, there will still be quite a drop in contract prices, with declines for both discrete and eMCP/uMCP product prices mostly falling between 10%-15%. Another thing worth observing is that mobile DRAM prices are usually negotiated by the quarter, demonstrating suppliers' resolve to deliver on shipments for the whole quarter. Thus current assessments give that the effects to arise from the Japan-South Korea incident will be limited in range, and will not pose too much of a problem to shipments hereafter.
Looking at price trends for mobile DRAMs 4Q this year, we see brands producing enough products ahead of January next year, in anticipation of the worker shortage to occur during the Chinese Lunar New Year, when people return to their hometowns. Total smartphone production is predicted to trend flat in 4Q compared to 3Q, standing at 360 million units. Falling mobile DRAM ASP will spur brands to raise memory capacity per phone, and will help speed up the clear-out process for suppliers. On the other hand, mobile DRAM ASP, having experienced three consecutive quarters of declines between 10%-15% beginning with 1Q this year, has fallen by 17% since it began rising back in 2016. This will test the limits of the top three suppliers' abilities to control fully-loaded costs, and suppliers are predicted to be relatively reluctant to lower prices in 4Q, leading to a narrowing of the decline.
As new production processes are being introduced, suppliers will be churning out mobile DRAM with faster transmission speeds and better-optimized power consumption. Since the LPDDR4 series may also be supported by low-mid range chips, its market share will be pushed up to 75 % this year. Adding the fact that suppliers are becoming passive with regards to LPDDR3 supply, we will see an accelerated phasing out of LPDDR3, whose market share (by bits) is forecast to drop below 15% in 2020. Although flagship phones implementing next generation LPDDR5 are poised to enter the market next year, we predict its penetration rate to lie below 10% in light of two considerations: its initial 20%-25% price difference with and its higher total production cost compared to the LPDDR4 series. The LPDDR4 series will remain market mainstream.