Key points of investment

Chalk announces 2024 interim results. 2024H1 chalk achieved revenue of 1.63 billion yuan, a year-on-year decrease of 3.1%; adjusted net profit of 0.349 billion yuan, an increase of 21.2% year-on-year.

The wrong exam schedule for public institutions disrupts revenue, and revenue is expected to remain flat throughout the year.

2024H1's training services achieved revenue of 1.379 billion yuan, a year-on-year decrease of 3.1%, mainly due to the collision between written examinations for public institutions and written examinations for provincial examinations, which led to a decline in course sales in public institutions. Among them, online learning products/large classes/small classes/special courses and promotion courses achieved revenue of 0.65/0.266/1.026/0.023 billion yuan respectively, accounting for about 4.0%/16.3%/62.9%/1.4% of total revenue. 24H1 book sales achieved revenue of 0.252 billion yuan, a year-on-year decrease of 3.0%, mainly due to damage to sales of books related to public institutions. We believe that by actively adopting strategies such as increasing the proportion of long-term classes and starting classes early in the second half of the year, the company is expected to narrow the revenue decline in public institutions in the second half of the year, and revenue for the whole year is expected to remain flat year on year.

Demand for public examinations has increased moderately, and the growth rate of small classes has verified product strength.

The number of jobs recruited for the 2024 national examination/provincial examination is yoy +6.6%/4.8%, while the number of applicants yoy is +13%/27% (there is a difference between the statistical caliber and the actual growth rate). The number of applicants is expected to increase moderately in the short to medium term. We expect revenue from 24H1 online learning products/large classes to be -12%/-28%, respectively, mainly due to market fragmentation and marginal decline in customer payment levels, and students' payment preferences are barreled. Large class classes show strong price sensitivity due to lack of after-school services. With the launch of AI teachers on August 1, the price of large class classes increased from 980 yuan to 1,280 yuan, an increase of about 31%. Large class products injected with AI testing services have obvious differentiation advantages, which can be expected to bring room to recover large class income; we expect 24H1 small class revenue to be +8% compared to the same period. It is probably mainly due to the fact that services account for a relatively high proportion of small class delivery, and the immediate requirements are more obvious. The company's product and service matrices are efficiently collaborated. As the OMO model continues to deepen, students quickly raise points, and the penetration rate of small classes with high customer unit prices is expected to continue to increase.

The gross profit margin reached a record high of 54.2%, and the average manpower efficiency of AI empowered teachers increased by about 9%.

2024H1's comprehensive gross profit margin was 54.2%, up 3.3 pct year over year. Among them, the gross margin of training services/book sales was 57.9%/33.6%, respectively, an increase of 4.1/-0.8 pct over the previous year. We believe that the main benefits are (1) continuous optimization of teachers' efficiency, with 2024H1 instructors contributing 0.518 million yuan per capita, an increase of 9% over the previous year; (2) the increase in the completion rate of small classes, and the company's OMO model continues to deepen, and the company's OMO model continues to deepen, setting up long-term classes in line with market demand while selling the short-term portion in a unitized manner. After the AI big model was applied, the average time for teachers to write reports was shortened from 20 minutes to 5 minutes. The AI face test saved teachers 50% of their time. Functions such as AI interview review and question submission and analysis were all online, with an average usage rate of about 90%. AI teachers are empowering people to improve efficiency and profit margins. Currently, one teacher can serve up to 150 people (100-120 in February '24, only 40 in February 2021).

The adjusted net profit margin exceeded expectations by 21.4%, and the three-fee forecast remained flat.

24H1 achieved an adjusted net profit margin of 21.4%, which surpassed expectations, with a year-on-year increase of 4.3 pcts, of which sales/management/R&D expenses ratio was 19.65%/12.11%/6.64%, +1.6/-6.9/-1.9 pct, respectively. The main reason was a strategic increase in marketing expenses and a reduction in SBC for administrative and R&D personnel. The company's employee structure was adjusted and optimized. As of the end of June 2024, the company had 7,564 employees, an increase of 239 over the end of 2023, which may be mainly due to an increase in sales, operations, personnel, etc. recruited to match the growth strategy; 3145 full-time teachers, 80 fewer than at the end of 2023, and the company's manpower efficiency continues to improve. The decrease in teachers is expected to be mainly due to strict assessment systems and normal loss.

AI teachers enhance user experience and teaching effectiveness, and the company pursues high quality, profit rather than revenue growth.

Starting August 1, the Chalk AI teacher “Chalk Head” will be officially launched. It can be used by public exam users who purchase the 25-season system class, online boutique class, OMO premium class, and test SVIP members. With the help of the prediction score function, the score increase for Premium Full Class/Premium System Class users reached 40%/16% (from 2022 to 2024 provincial joint testing). AI tools are being widely accepted and used by users. Chalk AI intelligent reviews are used more than 20 times more than manual reviews in interview scenarios, and are expected to bring about steady growth in market share and revenue scale in the medium to long term. As of August 9, 2024, the company has repurchased and cancelled 54.23 million common shares, with a cumulative repurchase amount of approximately HK$0.23 billion. Continued repurchases indicate the company's growing confidence. 2024H1's MAU is about 9.2 million, which is the same as the previous year, and has a large flow reservoir. We expect that leading companies will seek to improve product strength and reduce operating leverage, improve teaching quality and product profit margins, and not just seek rapid revenue growth.

Profit forecasting and valuation

As a leader in recruitment, testing and training, Chalk has shown strong product course development and technology research and development advantages in the current economic environment and industry competition. The launch of AI teachers on August 1 brought about an overall improvement in system class products and service capabilities, and the introduction of strategies such as superimposed unit sales for premium class leaders has increased the full class rate of small classes. As the OMO model continues to deepen and AI continues to improve efficiency under the advantages of online technology, there is still room for improvement in profit margins. We expect the company to achieve total revenue of 3.059/3.213/3.363 billion yuan in 2024-2026, up 1.24%/5.06%/4.65% year on year; adjusted net profit of 0.558/0.62/0.678 billion yuan, up 17.94%/11.10%/9.41% year over year. PE based on adjusted net profit to mother is 10.15/9.14/8.35X, respectively. We consider that short-term pressure on the company's revenue side comes from a marginal decline in the willingness of students to pay in low price courses and increased competition in the industry. It is expected that a series of positive measures and changes that the company will take in the second half of this year and next year will all be expected to bring about a recovery on the revenue side. It is recommended that the company continue to pay attention and give the company an “increase in weight” rating.

Risk warning

The increase in the number of people paying for courses and customer unit prices fell short of expectations; risk of fluctuating demand for recruitment examinations; risks such as changes in examination format, changes in the number of recruits, and changes in examination time due to changes in examination policies; increased risk due to industry competition, etc.