Breaking Down Shein's IPO into a Single Table

As someone who works with data archives, I recently broke down Shein's IPO into a table. Conclusion first: it's worth doing. Predicting stock price isn't the focus. I care more about breaking down emotional terms like "bleeding IPO" into verifiable fields: issue price, net proceeds, current valuation, peak valuation, old shareholder compensation, quarterly losses. Once you're done, you'll realize that judgments in headlines mostly come from a table that wasn't shown.

A common mistake is immediately searching for Shein IPO data. Listing fields first is more stable. I rebuilt a template in Excel over the last two weeks; the process is straightforward. Open Excel, create a new blank workbook. Write headers in the first row: Field, Value, Unit, Source, Notes. Start filling in Issue Price, Global Offering Shares, Net Proceeds, Current Valuation, Peak Valuation, Valuation Shrinkage Rate, Old Shareholder Compensation, Quarterly Losses in order starting from the second row. Leave a source next to each number. Shein listed on the HKEX on September 1st, with an issue price of HK$48.56, global offering of approx. 280 million shares, net proceeds of approx. HK$13.2 billion (equivalent to $1.74 billion USD). Current valuations vary by media and time point ($26.3B, $27B, $30B); I noted these discrepancies in the Remarks column. Use $98.2 billion for peak valuation; shrinkage rate calculates to approx. 73%. Write $3.5 billion for old shareholder compensation, $99 million for quarterly losses, and $395 million for net profit in the same period last year. One table contains both raw numbers and sources. Don't underestimate sources; verification relies entirely on them later.

After filling in raw numbers, add two formula columns. Assume current valuation is in B5, peak valuation in B6, enter =(B6-B5)/B6 in B7. Select B7, click Percentage Style in the Home tab; the number becomes a percentage format. If current valuation is $26.3 billion and peak valuation is $98.2 billion, the result should be close to 73%. Current valuation and issue price currencies differ; don't force calculate yet. Mixing HKD, USD, and CNY in one table easily biases conclusions. Add a Proceeds/Current Valuation row, entering =B4/B5. This ratio shows how heavy the money raised from the public market is relative to overall valuation. If news says "bleeding," add a column for "Signal Triggered." Fill "High" if valuation shrinkage exceeds 50% or if old shareholder compensation converted to USD exceeds net proceeds. My verification shows the most useful part is breaking "bleeding" into three questions: How much did old shareholders get back? How much did the public market raise? How far is the valuation from the peak? Headlines won't answer for you; tables will.

Errors happen most easily in three places. Numbers with "approx.," "about," or "hundred million" aren't recognized by formulas. Only fill in pure numbers, put "approx." in remarks. For example, fill 48.56 for issue price, unit as HKD. Inconsistent currencies: Net proceeds approx. HK$13.2 billion, equivalent to $1.74 billion USD; open a separate column for USD basis, convert manually first, then calculate ratios. Conflicting valuation bases: Pre-IPO reports said valuation might drop to $30 billion; post-listing there are claims of $27 billion and $26.3 billion. Don't rush to delete old numbers; keep timestamps.

I previously wrote about minimum rejection rates; the viewpoint is similar. If a metric is too coarse, it flattens complex issues. Capital news is the same; looking only at breaking below issue price or "bleeding" easily misses where the money comes from, who gets compensated, and why listing was necessary.

Next step, turn this table into an AI Company IPO Tracker. Change fields to Compute Power, Model, Customer Concentration, Regulation, Gross Margin. Then let ChatGPT or your usual model generate a summary based on the table, but you must make it list cited rows. Prompt can be written like this: Summarize only based on table content, do not supplement external facts, mark field source for each judgment.

This process isn't fancy, but it's durable. Next time you see a unicorn "bleeding IPO," you don't need to be led by the headline first; ask if they dare to lay out the fields.