Quick Pitch: Bel Fuse (BELFB)

Share Class Arb – 28% Upside

Pretty interesting share class arbitrage idea has just become public on VIC (free account is required). Below, I provide an updated snapshot of the situation together with some additional details.

Bel Fuse has two classes of common shares: Class A (BELFA) and Class B (BELFB). Over the last 10 years, class A used to mostly trade either in line with class B or at a slight discount. However, recently BELFA’s premium over BELFB has surged to around 30%. The investment thesis is simple: this premium is unsustainable and will eventually normalize.

Both share classes have equal economic value, but there are a few differences. Class A shares have voting rights, while Class B shares do not. Instead, BELFB holders receive dividends that must be “at least 5%” higher than those of BELFA. Here’s the overview of the share class structure and ownership:

It’s no surprise that on average BELFA traded in line or slightly below BELFB. The individual perks of each class (voting rights for class A and higher dividends for class B) are essentially insignificant. The dividends have historically been negligible, so the difference in value between the two classes should be minimal. Currently, BELFB’s dividend is $0.07/year, and BELFA’s – $0.06/share, which translates to annual yield of 0.35% and 0.24% respectively.

Meanwhile, the voting rights for BELFA come with a strange restrictive provision: if any shareholder increases ownership of Class A shares above 10%, it is required to also purchase enough Class B shares so that the total economic interest in Bel Fuse (Class B + Class A) matches Class A stake. Otherwise, all voting rights are automatically forfeited. This means that to obtain a meaningful portion of voting stock, investors are also forced to buy a very disproportionate amount of Class B shares (since there are far more of B shares outstanding). This condition pretty much eliminates the whole benefit of holding “the voting class”. For example, GAMCO (Gabelli Funds) owns 14% of class A stock, but doesn’t hold any class B shares. And a result, it can’t vote and will not be able to unless it reduces BELFA’s position below 10% or buys a ton of class B shares to comply with the condition. That’s quite unusual.

So, it’s quite strange that this year BELFA’s premium increased to 25-33%:

Reasons for this move are not exactly clear. However, it appears that the spread has materialized through three main spikes, as reflected in the graph above. Timing of these spikes closely match the following events:

• Nr. 1 – Q4 results release on February 21. There are two ways it might’ve impacted the spread. First, the earnings were quite poor, and both stocks dropped sharply. Since BELFB is the more liquid one (and has better price discovery), it simply dropped more, creating a premium for BELFA. The interesting thing, however, is that the premium hasn’t corrected soon after and even continued to increase slightly. So, probably the other reason had a bigger impact here. This other reason is that Bel Fuse launched its first-ever buyback program ($25m size) during the Q4 results release on February 21. Allocation of repurchases was initially structured based on the individual proportion of BELFA and BELFB to the total number of shares outstanding. The plan was to buy $4m of Class A and $21m of Class B. Since the free float and liquidity of Class A shares are way lower than Class B’s, the market may have started expecting that buybacks will give a disproportionate boost to BELFA price. That’s just my guess, but after this announcement, the spread immediately jumped from mid-to-high single digits to high teens.
• Nr. 2 – Russell Index Inclusion (May 24). Class A shares were added to the Russell preliminary additions list on May 24. In the following week, the spread surged from mid-teens to the high 20s.
• Nr. 3 – Q2 Results Release (press release on July 24 and full report on July 31). After the release of Q2 results, the spread went up from low 20s to 30%+. ‘m not really sure what triggered this move. Perhaps investors were pleased that management continued to execute on the buyback ($14m of stock was acquired in H1’24: $1.4m of Class A and $12.7m of class B), or maybe there was some delayed incremental buying pressure following the Russell inclusion in late June.
• Additionally – VIC author mentioned hearing rumors that some large new buyers had been building positions solely in Class A shares (as many funds can only purchase voting stock) since the beginning of the year. This might’ve added some additional buying pressure on BELFA price.

A long position in BELFB paired with a short in BELFA offers a way to profit from the expected convergence of the price gap. Keep in mind that BELFA’s borrow availability is somewhat volatile, fluctuating between 100k shares and 2-3k shares. The annual fee is at 6%. While there might be some hedging related risks here, I think those are quite low. Even though BELFA’s float and liquidity is much lower than BELFB’s, it the public float still stands at $100m. Moreover, the fee rate has historically been very stable. The risk of borrow completely disappearing and brokers being unable to find more shortable shares (leading to a potential forced buy-in) seems limited.

Similar share class arbs usually come with two drawbacks – uncertain timeline and no hard catalyst. This applies to Bel Fuse as well. It’s interesting that a similar spread previously appeared in 2022 and it took about a year for it to narrow down to average.

I haven’t been able to pinpoint the exact reason for the premium widening two years ago. However, it’s worth noting a couple of things: 1) stock prices were much lower back then (see graph below), so the actual dollar value of the gap was much smaller than it is now; and 2) both stocks were far less liquid than now (BELFA was especially illiquid), making it harder for arbitrage players to step in and help close the discount. This time around, the spread might close faster.

Last week, Bel Fuse announced a pretty sizable ($320m) acquisition of Enercon Technologies, expanding into the higher-margin Aerospace & Defense space. The market clearly loved this deal – BELFA shot up 13% and BELFB jumped 16%. It’s interesting is how this big liquidity event (3x-5x normal volume) shook up the BELFA/BELFB spread, narrowing it from 31% to the low 20s at one point the following day. While the premium bounced back eventually, it illustrates how fragile the price gap is and that it can tighten quickly when liquidity hits.

What’s also worth noting is that the end of March, shortly after the spread has appeared and widened to 20%+, GAMCO started selling down its Class A stake (April’s 13D/A, August’s 13D/A, September’s 13D/A) and has now reduced it from 17% to 13.6%. I can only speculate on the incentives here, but the bottom line is that if the current spread persists at this size, more and more investors will become willing to switch their Class A’s into Class Bs.

Overall, I find the setup quite compelling.

 

Business background

Bel Fuse is an industrial company that designs and manufactures electronic components, including fuses, power products, cable assemblies, and fiber connectors. While these products are largely commoditized, they are mission-critical for clients in industries like aerospace/defense, EV, telecom, and others. For these clients, such components are low-cost but carry high risk. Therefore, despite the commoditization, the margins in this space are quite solid, with operating margins ranging from the teens to low 20s.

BELFA has historically underperformed its much larger peers based on margins. However, things began to improve after the new CFO came on board in 2021 and started focusing not just on cost management but also on product management (which the company struggled with a lot previously). This shift has led to a successful business turnaround, which is clearly reflected both in margins and the share price.

At 12x NTM EBITDA, the stock appears more or less fairly valued now. The discount to peers (APH – 22x NTM EBITDA, TEL – 12x, CTS – 12x, LFUS – 15x) has narrowed substantially over the last few years, especially after this month’s rally. Remaining discount to higher valued peers might be justified by smaller scale, slightly lower margins, the unusual voting class provision (which doesn’t apply to management’s pre-IPO shares – 18% voting power), and historical issues with leverage (now resolved).

Pretty nice and concise business overview of Bel Fuse can be found in Atai Capital Management’s Q2’23 letter.