After buying $2K worth of Broadcom last week, I decided it would be fair to present my complete thesis and not just a few short fragments of it. So here it is: We all know a consensus management-led bull case stands in the expected $30 of EPS figure for FY 2028, which lands $AVGO just above $600 per share from here if we do the maths with a simple, index average 20x multiple. That's a double within 20-ish months FYI. Now, can Broadcom really generate $30 dollars of EPS by then - What specific numbers and financial assumptions make that possible? - At $30 of EPS, is a 20x TTM multiple reasonable or not? On their Q3'26 Earnings Call, management called for a total of $115B in 2027, and $230B in AI Semiconductor Revenue by FY2028 end. That is indeed two doubles in two fiscal years in a row! So let's dig deeper and make a PROPER THESIS; What we have now (point of relevance) - Q3'26 | NOTE: We will use most of those metrics as they are, implying ZERO GROWTH to make it a flat, simple base case. 1: Total revenue $29.59B +86% 2: Semi's solutions revenue $20.84B +127% 3:Software revenue $8.75B +29% 4: AI semi's revenue $16.7B +221% 5: Non-GAAP operating income $20.10B +92% 6: Non-GAAP net income $16.37B +95% 7: Non-GAAP diluted EPS $3.32 +97% 8: FCF $13.67B +95% Two things stand out to me; 1: $16.37B of non-GAAP net income in 1 quarter (equivalent to an annualized $65.5 billion if that quarter was repeated four times without growth) | 2: Profitability is keeping up with growth. Non-GAAP operating income grew 92% while revenue grew 86%. The company is not merely selling more chips; it is converting a large share of incremental revenue into operating profit. Now, let's split those numbers into 3 main revenue sources (2 and 3 are considered FLAT for 2 years for illustrative purposes) | 1: AI Semiconductors (accelerators, networking & optics) - $230B management's stated 2028 outlook | 2: Non-AI Semi's (storage and connectivity) - last reported quarterly RR of $4.3B | 3: Infrastructure Software (including VMware) - last reported quarterly RR of $8.7B On an annual basis we get: $230B + 4*$4.3B + 4*$8.7B = $282B for FY2028 if the business stays FLAT from here and the AI part grows as projected. Now let's convert this number into OPERATING INCOME; - Last reported non-GAAP op. margin is roughly 66%, resulting in operating income being $186B with our number. (keep in mind, we are acting like their margins won't grow during the 24-month period for illustrative purposes) Now, let's convert operating income to NET INCOME; - With todays Net Interest and other operating costs PLUS the taxes of 16%, we lose about $32B - landing just above $153B for clean, net profit. That's our figure of the day - $153B of net income after the company illustratively grew ZERO PERCENT for 8 QUARTERS and only got their AI bills paid by the customers who are locked-in with their NCNR contracts. The math is simple now; $153.4B devided by 4.9 billion of diluted outstanding shares equals to $32 of EPS. What's Broadcom's average 10-year TTM multiple on earnings? 42x to 54x 42 times $32 gets us to $1350 per share by the end of 2028. Keep in mind - this is me researching, basing my maths on HISTORY, CONFIRMED NUMBERS and KEEPING UN-CONFIRMED PARTS FLAT Matter of fact, we all know they are going to have much more than $30 of EPS because their other businesses are OBVIOUSLY going to grow as well. The bear case again? Meta, OpenAI, Anthropic and Google go to ZERO, reach financial INSOLVENCY and call it quits. The REAL case? They probably won't go anywhere and WILL pay Broadcom their lunch. NOT FINANCIAL ADVICE, DO YOUR OWN RESEARCH, DO YOUR OWN DILLIGENCE. Thank you for reading