Latest / Investor Exchange / Is PC Partner The Most Undervalued Way To Ride The NVIDIA Wave?

Transcript

Is PC Partner The Most Undervalued Way To Ride The NVIDIA Wave? · Investor Exchange

0:02
Welcome to The Debate. Today, we are examining a company that operates in one
0:08
of the most exciting yet notoriously volatile corners of the technology market.
0:14
PC Partner Group or PCPGSP.
0:18
They're the manufacturer behind some really iconic graphics card brands like
0:23
Zotac and Inno3D, operating squarely within the add-in board or AIB segment.
0:30
And the timing for this discussion is, well, it's critical.
0:33
We're coming off a fresh initiation report that just slapped an outperformed
0:37
rating on PC Partner with a pretty ambitious price target of SG 1.73.
0:43
This is all following a really a dramatic multi-year recovery since the whole
0:48
crypto crash and the inventory mess that followed.
0:51
Exactly. So the central investment question before us is this.
0:54
Given PC Partners' recent financial re-acceleration in 2024 and the first half
1:00
of 2025, you know, driven by this new product cycle and some strategic shifts.
1:05
Does the stock currently represent an undeniable buy opportunity,
1:10
one fueled by strong, visible earnings growth?
1:12
Or, and this is my view, is the company's deeply embedded history of volatility
1:18
and its fundamental reliance on a single major supplier, is that a major structural constraint?
1:24
Something that actually justifies the current low valuation,
1:27
making it a cautious hold or maybe even a sell for anyone looking beyond this current peak.
1:33
So I'll be presenting the skeptical case today, emphasizing that risk and that
1:38
dependency. And I will argue the bull case.
1:41
I believe the market is fundamentally mispricing the stock.
1:45
The deep discount is unwarranted given management strategic execution,
1:49
the robust recovery we're seeing, and, frankly, the clear catalyst for closing
1:53
that valuation gap with its peers.
1:56
The bull case for PC Partner is fundamentally a story of successful cyclical
2:00
execution and just compelling deep value.
2:03
The company has achieved what we'd call a clean re-acceleration,
2:07
powered primarily by the launch of NVIDIA's next-gen product,
2:10
the RTX 50 series or Blackwell.
2:12
And we're not just talking about forecasts here. The first half 2025 results are concrete proof.
2:18
They cleared over HK$6 billion in revenue, HK$6,355.3 million to be exact,
2:26
with really strong earnings per share or EPS of HK 0.645.
2:32
The numbers are strong, I agree. But look at the quality of that revenue.
2:36
Their own brand sales, you know, Zotac and NO3D, jumped over 60% year over year.
2:43
Now, this is crucial because it confirms their strong relationship with NVIDIA.
2:47
And that means they've secured access to the highest margin,
2:51
top-tier SQs, the stockkeeping units, like the flagship RTX 5090.
2:56
That means better margins, not just more volume. And what's more,
3:00
management is structurally de-risking the business.
3:03
They implemented the China Plus One strategy, setting up their headquarters
3:06
in Singapore and ramping up a plant in Bataan, Indonesia.
3:10
And this proved its value almost immediately, letting them reroute U.S.-bound
3:14
assembly seamlessly after the February 2025 tariffs.
3:18
That reduces the kind of policy friction that plagues their competitors.
3:21
I come at it from a different perspective.
3:24
While the immediate results are impressive, we just cannot ignore that PC Partners'
3:28
entire story is defined by extreme, almost traumatic cyclicality.
3:33
It just makes this current high-growth phase inherently suspect in the long term.
3:38
This isn't just a business that rides general economic cycles.
3:41
It's an AB manufacturer caught in these hypercycles of GPU mining booms and
3:46
the subsequent brutal crashes.
3:48
But they survived the downturn, and they retained their net cash position.
3:51
But at what cost to investors?
3:53
We saw net income balloon to a massive HK$2.37 billion during the 2021 mining
4:00
peak, only to crash down to a mere HK$60 million in 2023.
4:05
That's a 97% decline.
4:08
I mean, investors have very long memories when cycles are that brutal,
4:11
and frankly, the growth trajectory is already showing signs of a quick peak.
4:16
The forward forecasts already project a significant revenue decline post-2025,
4:22
dropping from HK$12.72 billion next year down to HK$10.78 billion in 2026.
4:30
This isn't long-term sustainable growth. It's a temporary bump.
4:34
But that's precisely why the valuation is so absurd.
4:38
The valuation is low because the fundamental risk is structural.
4:41
It's single-supplier dependence. PC Partner relies almost entirely on NVIDIA's
4:47
GPU technology, which, you know, dominates 92% of the discrete desktop GPU market.
4:53
This means NVIDIA holds the lever on capacity allocation, effectively capping
4:57
PC partners' upside and its volume predictability.
5:00
And finally, the nature of the AIB business just guarantees low long-term profitability.
5:05
Even in this recovery, the net profit margin for 2024 was only 2.6%.
5:09
They are a high-volume, low-margin partner.
5:12
They are not a technology innovator. I agree that historical volatility is a
5:17
constraint. But I think you're underestimating the factors that differentiate
5:21
this recovery from the 2021 crypto boom.
5:25
This period is supported by multiple factors beyond a simple volume surge.
5:29
It's fundamentally about an improved product mix.
5:33
Securing those high-end RTX 5090 allocations means the average selling price,
5:39
the ASP, of their boards goes up.
5:41
And that drives margin expansion despite that thin 2.6% net margin you just
5:46
cited. And how much margin expansion are we actually seeing here?
5:49
Enough to make a real difference.
5:51
We are forecasting 2025 earnings per share, or EPS, to roughly double 2024's result.
5:58
So jumping from HK 67 cents to HK $1.04.
6:04
This is a dramatic recovery in core profitability.
6:08
And more importantly, management is looking strategically to smooth out that cyclicality.
6:14
They're expanding into genuinely adjacent, non-cyclical segments.
6:18
I'm referring to their push into GPU servers for AI, which gives them exposure
6:23
to enterprise compute and their existing electronics manufacturing services,
6:27
which offers a stable, reliable revenue base, even if it is lower margin.
6:32
See, I'm just not convinced that a temporary spike in profitability equates
6:37
to sustainable value, especially when the cycle is explicitly forecast to turn
6:41
so quickly. Let's focus on the forecasts themselves.
6:45
You cite the doubling of EPS to HK$104 in 2025, but the very analysts projecting
6:50
that strength are also forecasting a sharp decline in 2026.
6:55
EPS growth is projected to fall by 22.3%, right back down to HK$.81.
7:01
That indicates the Blackwell cycle is expected to peak quickly in 2025,
7:04
which mirrors exactly the kind of leverage reversal we saw in prior cycles.
7:08
But even at HK$.81, the company is still dramatically undervalued.
7:14
Perhaps, but my skepticism grows when we consider the structural pressure coming from the cost side.
7:21
Industry reports are already signaling rising memory inflation.
7:24
We're talking tightness in DRAM and especially GDDR7 pricing through 2025 and 2026.
7:31
As an AIB manufacturer, PC Partner is buying these crucial expensive components.
7:37
When that cost stack lives, it puts immense pressure on gross margins,
7:41
especially when you're competing in high-volume, price-sensitive markets.
7:45
This component inflation is a very real cap on the sustained profitability you're projecting.
7:50
What margin assumption is even underpinning your HK$1.04 EPS projection?
7:56
Are we truly factoring in GDDR7 pricing, or is that just based on volume?
8:00
Okay, that's a fair challenge regarding component costs. They're definitely a headwind.
8:05
However, I would frame the operational de-risking strategy differently.
8:10
I see it as a structural hedge against these external pressures.
8:13
The China plus one move is providing tangible policy friction reduction and
8:18
a competitive advantage right now.
8:20
The swift pivot to the bottom Indonesia facility to handle U.S.
8:23
Tariffs earlier this year is a testament to a newfound operational agility.
8:27
This ability to pivot reduces the cost of disruption, and it makes the business
8:31
inherently more robust than it was during that severe 2023 crash when they were
8:35
geographically constrained.
8:37
Agility is good, but let's talk about real impact.
8:40
Does this fundamentally alter the business model?
8:43
I mean, the documents state that 75% of revenue still comes from VGA cards.
8:48
The business is overwhelmingly cyclical and exposed to NVIDIA's product calendar.
8:53
If NVIDIA pushes a launch back by six months, PC partners sees a massive revenue
8:57
hole, regardless of where they happen to assemble the wards.
9:00
But that agility allows them to pursue diversification.
9:04
New products like the GPU servers, handheld PCs, these are small contributors today, I grant you.
9:10
But they signal a real commitment to moving revenue away from that core cyclical
9:15
gaming segment. This incremental diversification slowly builds intrinsic value.
9:20
The market isn't discounting these new projects because they don't exist.
9:24
It's discounting them because the documents themselves confirm they won't contribute
9:27
meaningful revenue until fiscal year 2027.
9:31
So for the immediate investment horizon, the company is still a GPU AIB partner.
9:36
And that means it's subject to the single source risk of NVIDIA and fierce competition
9:40
from companies like Asus and
9:41
MSI, who are much larger and have vastly more diversified revenue bases.
9:46
You know, everything from laptops to motherboards?
9:48
Until 2027, the diversification story is aspirational.
9:52
It's not a fact for the P&L. Right, but we have to pivot back to the valuation
9:56
because even acknowledging the risks you cite, The stock's valuation is just
10:00
too compelling to ignore.
10:03
It is trading at 7.5 times trailing price-to-earnings P.E.
10:06
And it's priced below book value, about 0.8 times price-to-book on the Hong Kong listing.
10:11
This just indicates very limited downside risk, assuming they merely execute
10:16
this cycle effectively.
10:17
Limited downside risk is not the same thing as guaranteed upside.
10:21
But the upside is clear. Our target price of SG$173 implies a substantial 70%
10:28
upside from current levels.
10:30
And crucially, this target price still only equates to a modest 8 times our 2026 projected EPS.
10:38
When you look at Taiwanese peers like Gigabyte trading around four times price
10:43
to book, PC Partners' valuation gap is just clearly unwarranted given its robust
10:48
net cash balance sheet and its recovering earnings.
10:50
It really suggests the market is pricing in the worst possible cyclical outcome,
10:55
and any deviation towards stability should trigger a massive re-rating.
10:59
I see why you feel that way. But let me give you a perspective focused purely
11:03
on the mechanics of that SG$1.73 target price, because I think it reveals deep market skepticism.
11:10
The analyst arrived at that number using a blended valuation methodology,
11:14
combining two key models.
11:16
A standard practice for volatile companies, sure. Exactly.
11:20
So while the discounted cash flow, or DCF model, which is what captures that
11:25
long-term growth potential,
11:27
suggests a theoretical upside of SG$3.39 per share, the analyst heavily weighted
11:32
the Conservative Dividend Discount Model, or DDM, assigning it a 70% weight.
11:37
And the DDM is necessary because it grounds the forecast in near-term shareholder
11:41
payouts, not speculative growth assumptions.
11:44
Precisely. And what does that DDM yield? A target price of only SG099,
11:51
which is right around the current market price of SG102.
11:54
This heavy weighting on the DDM, to me, signals that the market,
11:59
and arguably the analysts themselves, are anchored to the company's near-term
12:03
yield and remain deeply skeptical about long-term sustainable cash flow,
12:07
because of the volatility we've discussed.
12:10
The bullish SG-173 target is therefore fundamentally built on the low-weighted,
12:16
speculative upside from the DCF, not the guaranteed execution signaled by the DDM.
12:21
Why should we invest based on the analyst's high-risk assumption when they themselves
12:25
are weighting the low-risk ones so heavily?
12:27
The weighting simply reflects the inherent difficulty in forecasting AIB businesses.
12:33
However, my core argument holds, PC Partner is executing well into the new Blackwell
12:38
cycle. They are regaining high-end market access, and they are strategically
12:41
de-risking operations through China Plus One.
12:44
The stock's significant discount to its peers, coupled with a recovering earnings
12:48
trajectory, we're forecasting a strong 12.9% return on equity for 2025.
12:53
It just makes it a compelling buy.
12:56
The belief here is that this new operational resilience will narrow that staggering
13:00
valuation gap, making this a classic deep value cyclical play.
13:04
I maintain that the volatility risk and the overriding dependence on NVIDIA
13:10
are structural realities that fundamentally justify the current low valuation.
13:15
While the short-term cycle is overwhelmingly positive for 2025,
13:19
the long-term outlook remains structurally challenging.
13:22
You have inevitable downturns, and the rising component cost pressures,
13:25
like GDDR7, that will compress margins.
13:28
Investors should treat this stock as a cyclical hold, acknowledge the high dividend
13:32
yield during the good years, but remain extremely cautious about any dramatic
13:36
multiple expansion until that sustainable, non-cyclical diversification,
13:40
specifically the AI server projects,
13:42
actually materializes and meaningfully contributes to the P&L post-2027.
13:47
Two highly contrasting views on whether to chase this momentum or respect the cycle. Indeed.
13:52
The data offers strong evidence for both caution and for aggressive investment.
13:56
Before we conclude, we must state clearly that this discussion is for educational
14:00
purposes only and does not constitute financial advice.
14:05
Investors should always conduct their own thorough due diligence and assessment
14:09
of their own risk tolerance.
14:12
Absolutely. Thank you for joining us.