Ben Brostoff

Thank you @MSmicrocaps for giving me the opp to present $AMNF at the MSM Summer Virtual Conference. Deck here drive.google.com/file/d/1Hk9… . Armanino is the rare microcap with sane capital allocation, capex lite biz, #1 player in great market and growth runway. Frozen pesto bull!

$95bn on $META dilution ($12.5 on convertible $METAM and up to $82.5 on common) is about what my model had and based on early reaction seems to be better than feared. I do want to understand more why equity over debt - release mentions "diversify funding sources" and AA- rating.

Will answer my own question here with Pershing Q2 letter from 8/12 w/ table below. You need a little multiple expansion on 20% IRRs bc EPS growth <20% on all three. Overall though I like that Bill is choosing names on avg w/ faster EPS growth + lower starting multiple than S&P.

Looking at @BillAckman 's investing principles, recent interview and $META Q3, I think it's likely Bill believes equity raise unlikely, $META will generate substantial FCF regardless of next 3-5 years of capex spend and that Muse will lead to substantial long-term growth.

$TNIE competitor Yoto even though private does have history of annual reports - while Yoto is growing much faster, the gap is shrinking as Yoto decelerates and Tonies as of 25 end is 4x bigger on sales. 25' was a deceleration for Yoto from 86% to 43% y/y growth in USD while Tonies went from 33% to 31%. I make this point bc since I wrote Tonies up in August, most of the pushback has been on Yoto gaining share and anecdotally being more popular among older kids in the US. I think Tonies' size will continue to give it a substantial advantage and also I'd bet it will show operating leverage and be able to fund out of cash flow sooner than Yoto.

IMO $BRK.A should look at $CRWV debt. Very attractive yield and a lot of equity buffer. In @lloydblankfein 's book he talks about how $GS needed Buffett blessing to push down future cost of capital and calm market. $CRWV may be fine but optics matter, could be win-win for both.

Stock got bid today on Q3 update that orders are up 50% y/y + first wins in Mexico, Costa Rica and Slovenia. This is acceleration from Q3 25' growth (+23%). IMO what is happening here is CNS went from 1 to 6 products and AI made those products more important. $DDOSF $CNS.L

Per Zuck, it's $15 bn to "operate" a 1gw data center cluster per year after 50-$60 bn of upfront capex (I am guessing this includes depreciation). AI boom really is changing long term cost structure of $META and others. @alexandr_wang calls it "entrepreneurship on hard mode"

YTD portfolio reflections: - Leaders in growing markets have worked ($TWLO , $BLZE , $CRWD / $PANW / cybersec basket) - Got some luck on my side with outcomes on $MGM , $EWCZ, $CCLDO - Tax loss harvested $WSC and $EIC - levered cyclical + small CEF lessons learned

On $MGM , drawdown I think is deserved based on this data, rates, $PPLI bid falling apart and subsequent interest in $PPLI . No idea when bottom will be in on RevPAR. Wonder what we'll see on buybacks given they bought all through 30s, avg last 5y at $38.52

Tonies sold out Disney Princess bundle on $AMZN (since have restocked) and I have been seeing Pokemon set ads and deals for TB2 bundles, also $WMT has TB2 bundle sale for $109 versus $170 non-sale. I think Tonies is being aggressive growing install base into holiday season

Every time I see $BFAM in ValueLine I'm surprised by how it has compounded earnings + improved ROE v stock price. Will have doubled EPS in 4 years while multiple shrank from 30s to 12x. Enrollment, occupancy + quality concerns, but still think this one is interesting.

The Combs - Munger interview just shared on @shaneparrish pod echoes this point. Price can make great businesses bad investments. Ian writes in his book that most microcaps end up being <36m holding periods, and IMO changing conditions / price make it hard to hold longer.

Looks like $NET can block Dots from accessing content (interesting this does not happen with Muse). Even when I take over the browser control, it gets stuck in a verification loop. Will be interesting to see long term what happens with agents accessing websites from VMs...