Qingmei Group Holdings

[mrEngineer]

Hi d.o.g, I did not study the IPO prospectus or do not know the industry at all. But did you take a look at the CAPEX and whether PPE has increased during that period of time? It is possible that the machines are not Superman or Wonder woman but Iron Man. haha.

Another reason I can think of is that during FY07, Qingmei was not able secure enough contracts or work done and capacity utilization might be running below 70% while in FY09 it has risen to its peak? Chinese workers maybe cheap still in 2007 and the management do not mind hiring more people.

[weijian]

Most importantly from my one-to-one with Mr. Tan Siok Sing was the assurance that the Audit Committee (AC) has taken all pains to make sure that Qingmei does not turn out to be a fraud case like numerous other S-Chips. I had a similar assurance from Mr. Pek Yew Chai (Independent Director & Audit Committee member), who is a former CEO of Singapore Computer Systems (SCS). These clarifications offer a certain degree of assurance to minority shareholders like myself that we are not being taken for a ride.

I would really appreciate if any other members who had been to the AGM can pen down their thoughts as well.

hi sgmystique,
The first time i heard something similar at a S chip AGM, i was taken in (the company bombed 1 yr later and the IDs resigned). That is also the last time i would ever believe in such cr*p or attempt to seek such consolation..

As a matter of fact, i would implore u to challenge the IDs to put their $ where their mouth is, ie. use their director fees to buy up supposingly undervalued shares on the open market!

[momoeagle]

Now i have to disagree with you on the wonder woman part.
She cant be working in a shoe factory.
We all know she works for that undergarment factory.(think wonder bra)

Well, Wonder Woman could be working in the shoe factory as a sideline

[d.o.g.]

But did you take a look at the CAPEX and whether PPE has increased during that period of time? It is possible that the machines are not Superman or Wonder woman but Iron Man. haha.

Capex:
Instead of just taking capex from the cash flow statement, we need to see what was ACTUALLY added to plant and machinery, which we find on page A-27 of the IPO Prospectus, under additions to plant and machinery:

FY07: 74.3m
FY08: 20.4m
FY09: 17.5m

Plant and Machinery at year end:
FY07: 60.1m
FY08: 62.3m
FY09: 58.2m

1. Plant and Machinery actually declined slightly each year

The amounts added did not cover depreciation, which means the company did not add meaningfully to the productions lines that were already in place i.e. there could not have been any meaningful increase in actual capacity.

One explanation: the manufacturing process is not automated and highly labour-intensive. Therefore expanded capacity should be reflected in headcount rather than plant and machinery. However, the IPO prospectus itself also makes clear that there was no meaningful increase in headcount either.

So, very little invested in new machines, and very few new people hired. Yet, capacity and output increased dramatically over the 3 years. How do we solve this mystery?

Either they did hire Superman and Wonder Woman, or they were previously using old scissors and blunt knives, and suddenly they upgraded to advanced robots that someone threw out in the garbage. Maybe they found some wounded Transformers and nursed them back to health, and in return they agreed to be Qingmei's factory slaves? After all, there is surely "more than meets the eye" here.

So: superheroes working cheaply, or super robots rescued from the dumpster? Take your pick.

2. Numbers don't add up

Let's look at reported depreciation:

FY07: 19.8m
FY08: 26.0m
FY09: 29.2m

Now, we know that total PPE at any point consists of beginning PPE minus depreciation, plus capex.

So let's look at FY08.

Expected PPE at end of FY08
= 199.3m (start of FY08) - 26.0m (depreciation) + 52.5m (PPE added)
= 225.8m

The FY08 accounts show 196.2m. Where did the 29.6m go?

Now let's try FY09.

Expected PPE at end of FY09
= 196.2m (start of FY08) - 29.2m (depreciation) + 53.3m (PPE added)
= 220.3m

The FY09 accounts show 188.4m. Where did the 31.9m go?

So in the space of 3 years 61.5m simply disappeared from the balance sheet. Since the independent directors are well educated and highly experienced, I'm sure they must have taken a look at this and figured it out. I would love to hear their answer.

Another reason I can think of is that during FY07, Qingmei was not able secure enough contracts or work done and capacity utilization might be running below 70% while in FY09 it has risen to its peak? Chinese workers maybe cheap still in 2007 and the management do not mind hiring more people.

Let's look at the background of the Su brothers:

Su Qingyuan:
- 1990 production supervisor at Sucuo Packaging Factory
- 1996 factory manager at Jinjiang City Chendai Sucuo Qingmei Colour Printing
- 1998 trading sports shoe soles and their components
- 2003 started Fujian Qingmei.

Su Qingjiang
- 1987 publication printing operator at Sucuo Packaging Factory
- 1996 in charge of production operations and supervision of production team at Qingmei Colour Printing
- 1999 trading shoe materials
- 2003 joined Fujian Qingmei

Su Shubiao
- 1989 full-time executive member of the Sucuo Village Committee
- 1999 finance and administrative manager at Quan Xing Shoe Plastic Co
- 2003 joined Fujian Qingmei

Clearly, all 3 brothers had plenty of production and management experience by 2003, when they started Fujian Qingmei. Page 120 of the IPO prospectus explicitly says the company has "a strong management team" with "strong industry experience coupled with dedication and diligence". Should we then believe that in 2007, their 4th year of successful operations, they suddenly decided to hire extra people to sit around and drink tea?

Let's look at the RB Outsoles and Shoe Soles Capacity:
FY07: 27.8m
FY08: 42.6m
FY09: 45.6m

Actual output:
FY07: 11.8m
FY08: 23.9m
FY09: 30.9m

Implied Utilisation rates:
FY07: 42%
FY08: 56%
FY09: 68%

This is striking.

a. This is the first time I have seen a manufacturing company report 16.0% net margins while operating at only 42% utilisation (FY07). And at 56% utilisation, net margins reached 21.4% (FY08). This is truly an amazing company. But then, with Superman on the payroll, it should be.

b. Strangely, moving from 56% to 68% utilisation did NOT improve margins meaningfully. FY09 margins were 21.9%, not much better than FY08 despite the big jump in utilisation. I wonder why.

c. Why did the company bother to invest in PPE and increase capacity in FY08 and FY09 when utilisation the previous year was only 42% and 56% respectively? Normally, manufacturers only invest when utilisation hits 75-80%, some require 90% or even higher. Seasonal businesses have more slack, but the IPO prospectus says the business is not seasonal...

What an amazing company, indeed. The financial results are unbelievably good. No, wait. Maybe, instead of Superman and Wonder Woman, they've hired The Incredibles? That would literally make Qingmei an incredible company.

The usual disclaimers apply. I would be happy to see the above analysis countered with supporting data.

[donmihaihai]

The usual disclaimers apply. I would be happy to see the above analysis countered with supporting data.

Let me try to balance the whole view a little as its swing too far to an extreme.

1 ) Superman, Wonder Woman and incredible

IPO prospectus page 62.

“For FY2007, we only commenced sales in October 2006 and recorded sales for a period of nine months as compared to 12 months in FY2008.”

IPO prospectus page A-27

PPE at beginning of the year (FY2007)- nil
Addition - 74m
Depreciation - 14m
PPE at end of the year - 60m

PPE at beginning of the year (FY2008)- 60m
Addition - 20m
Depreciation - 18m
PPE at end of the year - 62m

PPE at beginning of the year (FY2009)- 62m
Addition - 17m
Depreciation - 21m
PPE at end of the year - 58m

Depreciation period of PPE - 1-10 yrs (page A-16)

Actual depreciation per machine is unknown but average depreciation can be re-calculated. But that is not my concern.
My concern is PPE and depreciation does support sale for FY2007 was not for 12 months. Add in production start up issue, learning curve, etc, it does not support superman and wonder woman thing. Maybe someone with knowledge can point out that it is impossible to has 8,590 pair of shoes per worker per year.


2) Utilisation rate and gross margin.

Note : gross margin is used rather than net margin because in production, overhead expense goes to direct cost and admin, selling and distribution expenses is minimum for this company.

IPO page 59

Direct material as % of COS - FY2007 - 62.8%, FY2008 - 65.4%, FY2009 - 65.4%
Manufacturing overhead as % of COS - FY2007 - 14.9%, FY2008 - 12.1%, FY2009 - 12.7%
Direct labour as % of COS - FY2007 - 19.4%, FY2008 - 19.3%, FY2009 - 19.7%

Explanation of Direct material, manufacturing overhead & direct labour can be read on the same page.

My point is when manufacturing overhead is as low as 12 - 15%, utilisation rate is not the key to margin. Direct material is.

3) Disappearing of PPE from B/S

IPO prospectus page A-4
Other payable and accrual - 149m, 144m, 48m
IPO prospectus page A-6
Increase in other payable and accrual - 16m, 7m , 18m
Payment for purchase of PPE - (130m), (52m), (53m)

IPO prospectus page A-27
Addition of PPE in note 13

IPO prospectus page A-32
Other payable and accrual in note 20

Please do the calculation.

Note : A company can cooked it book but it can’t make it disappear just like that. (I have done the same thing for Ezra 2 or 3 years back but that was just for cash flow statement for financial statements but not a full report. There is a different and for cash flow statement, it can be present in all kind of way and even omit information but still look good. But that is for cash flow statement only)

NOTE : I know nothing about this company BUT I don’t think they can cook it book in such simple way.

[D123]

Now, we know that total PPE at any point consists of beginning PPE minus depreciation, plus capex.

Just to add to this, I think movements in the PPE account can also be affected by disposals of PPE, as well as transfers to other accounts. So the formula should also include "minus disposals of PPE and transfers to other accounts".

(I should add too that for Qingmei, there were no disposals in the 3 years 2007, 2008 and 2009, and the only transfer of account items happened in 2007 from a sub-PPE account to another sub-PPE account.)

[mrEngineer]

[2. Numbers don't add up

Let's look at reported depreciation:

FY07: 19.8m
FY08: 26.0m
FY09: 29.2m

Now, we know that total PPE at any point consists of beginning PPE minus depreciation, plus capex.

So let's look at FY08.

Expected PPE at end of FY08
= 199.3m (start of FY08) - 26.0m (depreciation) + 52.5m (PPE added)
= 225.8m

The FY08 accounts show 196.2m. Where did the 29.6m go?

Now let's try FY09.

Expected PPE at end of FY09
= 196.2m (start of FY08) - 29.2m (depreciation) + 53.3m (PPE added)
= 220.3m

The FY09 accounts show 188.4m. Where did the 31.9m go?

So in the space of 3 years 61.5m simply disappeared from the balance sheet. Since the independent directors are well educated and highly experienced, I'm sure they must have taken a look at this and figured it out. I would love to hear their answer.

I was not interested in this company but since there may be some discrepancies, it may worth some time to look into it.

The problem may lie in the way you have calculated additions to PPE. Using your FY08 as an example,

Expected PPE at end of FY08
= 199.3m (from A-27) - 26.0m (from A-27) + 52.5m (from A-6 CF statement)
= 225.8m

The question is whether is it fair to use 52.5m from cashflow statement as addition to PPE. If we look back to A-27 for FY08, the stated addition was 22.9m. So if you take 52.5m - 22.9m, there you get your missing 29.6m.

Why does this happen? I guess this is because the cash outflow CFI for PPE in FY08 has some amounts relating back to FY07. FY07 addition of PPE was 191.2m (A-27) and compare to FY07 CFI cash outflow for PPE was 129.6m (A-6). This means that in FY07, QM has recognize 191.2m of PPE but only paid for 129.6m cash and has the remaining amounts to pay for the next following years.

Do correct me if I am wrong. I am always willing to learn.

---

2) Utilisation rate and gross margin.

Note : gross margin is used rather than net margin because in production, overhead expense goes to direct cost and admin, selling and distribution expenses is minimum for this company.

IPO page 59

Direct material as % of COS - FY2007 - 62.8%, FY2008 - 65.4%, FY2009 - 65.4%
Manufacturing overhead as % of COS - FY2007 - 14.9%, FY2008 - 12.1%, FY2009 - 12.7%
Direct labour as % of COS - FY2007 - 19.4%, FY2008 - 19.3%, FY2009 - 19.7%

Explanation of Direct material, manufacturing overhead & direct labour can be read on the same page.

My point is when manufacturing overhead is as low as 12 - 15%, utilisation rate is not the key to margin. Direct material is.

Interesting statement on the utilisation rate and margin. Learnt something from here. But, in my view, the low manufacturing overhead margin does not correlate to the need for investment. It might be possible that QM clinched a new customer (e.g. Nike or ASICS) and had to increase their production capacity or maybe even their production technology in order to meet the demand.

d.o.g is right on the capacity portion as it is puzzling for me for such low utilization rates to increase PPE unless there are synergies for the new PPE addition in terms of transportation networks or location of the new plants where they can increase their demand volumes and prices. I can imagine the ROCE figures for the new additions to be incredible as well.

[d.o.g.]

Maybe someone with knowledge can point out that it is impossible to has 8,590 pair of shoes per worker per year.

Qingmei only makes the soles. The soles are formed by:

EVA I Midsoles: plastic injection moulding
EVA II Midsoles: foam injection moulding
RB Soles: stamping of synthetic rubber slabs

These are all machine-dependent processes.

Cycle time can be estimated given the hours, output and number of lines.

The prospectus has a full breakdown on page 106. They calculate capacity based on 23 hours/day, 29 days/month, and the actual number of months the machines were available.

Let's look at FY07:

EVA I:
===
Capacity = 15,306,000 pairs

Machine-hours available
= [(7 lines * 9 mths) + (2 lines * 3 mths)] * 29 days/mth * 23 hours/day
= 46,023 hours

Cycle Time per machine = 0.00301 hours per pair, or 10.8 seconds per pair
Output per hour per machine = 332.6 pairs per hour

How many workers would you need?

There is one new pair of soles to handle every 11s, if there is one worker working on the same cycle time then you only need one worker. But this goes on 23 hours/day, 29 days/month. So more realistically you need at least 3 workers, assuming each works ~8 hours/day. Suppose you decide to add some slack to allow workers to only handle soles every minute since they have to cut/trim the soles, so you need 5.5x the number of workers, or 16.5. Call it 17 workers. Let's add 6% more workers to account for the 3 golden week holidays (3 weeks out of 52). So we need 18 workers per machine, and 9 machines, to produce 15.3m pairs of shoes over 9 months (weighted for actual availability). Total is 162 workers.

EVA II Stage 1
===
Capacity = 17,290,000 pairs

Machine-hours available
= 3 lines * 9 mths * 23 hrs/day * 29 days/mth
= 18,009 hrs

Cycle Time per machine = 0.00104 hours/pair, 3.75 seconds per pair
Output per hour per machine = 960.1 pairs per hour

We need about 3x as many workers per machine as for EVA I i.e. 54 workers per machine, for 3 machines = 162 workers.

EVA II Stage 2
===
Capacity = 13,833,000 pairs

Machine-hours available
= [(15 lines * 9 mths) + (4 lines * 3 mths)] * 29 days/mth * 23 hours/day
= 98,049 hours

Cycle Time per machine = 0.00709 hours/pair, 25.5 seconds per pair
Output per hour per machine = 141.1 pairs

We need 40% as many workers per machine as for EVA I, so 7.2 workers per machine, for 19 machines, total 137 workers.

RB outsoles and shoe soles
===
Capacity = 27,813,000 pairs

Machine-hours available
= [(13 lines * 9 mths) + (3 lines * 3 mths)] * 29 days/mth * 23 hours/day
= 84,042 hours

Cycle Time per machine = 0.00302 hours/pair, 10.9 seconds per pair
Output per hour per machine = 330.9 pairs

We need about the same number per machine as for EVA I i.e. 18 workers per line, for 16 lines, total 288 workers.

This gives us a total production workforce of 162 + 162 + 137 + 288 = 749 workers

For FY07 Qingmei reported 2,986 workers. So we are off by a factor of 4x. This implies that on average it takes 4 min (as opposed to initial estimate of 1 min) for a worker to process 1 pair of soles. On the other hand, by the company's own figures, it only takes the machines 3.75-25.5s to make one pair of soles, inclusive of loading, injection, curing, unloading and cooling times.

Remember that the soles are made in large batches, especially the rubber soles which are just stamped from sheets of rubber. So one worker handling one sheet of rubber would effectively be moving dozens of soles at a time. Likewise for plastic/foam injection moulding where you can mould 3-4 pieces at a time.

Try to picture what this means on the factory floor:

The machines a new batch of soles every few minutes, each batch consisting of several pairs of soles (effective cycle time per pair is 3.75-25.5s). Workers pick up the batch for processing. Every 4 min, a worker finishes processing one pair of soles (cutting, trimming etc). The implication is that the manufacturing process is highly labour-intensive, with far more time spent on human processing than on machine manufacturing. Therefore, to increase output, it is more important to increase the workforce than to increase machine capacity.

This brings us back to the issue of manpower. Big changes in output, yet there was no change in value of plant/equipment and no meaningful change in manpower.

3) Disappearing of PPE from B/S

IPO prospectus page A-4
Other payable and accrual - 149m, 144m, 48m
IPO prospectus page A-6
Increase in other payable and accrual - 16m, 7m , 18m
Payment for purchase of PPE - (130m), (52m), (53m)

IPO prospectus page A-27
Addition of PPE in note 13

IPO prospectus page A-32
Other payable and accrual in note 20

Please do the calculation.

Thanks. Yes, problem solved. The company recognized 191m of PPE in FY07 but only paid 130m. The shortfall was made up in FY08 and FY09. Most of it was the factory (construction cost).

My point is when manufacturing overhead is as low as 12 - 15%, utilisation rate is not the key to margin. Direct material is.

I did a quick check in my master spreadsheet. Almost all the manufacturing companies in my investment universe charge less than 5% of sales to depreciation and amortisation, it seems to hover around 2-4%. For FY07-FY09 Qingmei charged 7.3%, 4.4% and 3.7%, not unusual.

Every manufacturing company I have ever talked to (which must be in the dozens by now) is focused on utilisation (which they can control) and not raw materials (which they cannot control).

Like I said, this is the first manufacturing company I have ever seen that can book a 16% net margin at only 42% utilisation, and 21% net margin at 56% utilisation.

As usual, YMMV.

The problem may lie in the way you have calculated additions to PPE. Using your FY08 as an example,

Expected PPE at end of FY08
= 199.3m (from A-27) - 26.0m (from A-27) + 52.5m (from A-6 CF statement)
= 225.8m

The question is whether is it fair to use 52.5m from cashflow statement as addition to PPE. If we look back to A-27 for FY08, the stated addition was 22.9m. So if you take 52.5m - 22.9m, there you get your missing 29.6m.

Why does this happen? I guess this is because the cash outflow CFI for PPE in FY08 has some amounts relating back to FY07. FY07 addition of PPE was 191.2m (A-27) and compare to FY07 CFI cash outflow for PPE was 129.6m (A-6). This means that in FY07, QM has recognize 191.2m of PPE but only paid for 129.6m cash and has the remaining amounts to pay for the next following years.

Yes, mystery solved. The company got PPE in advance in FY07 and paid up in later years.

[Big Toe]

Good analysis as usual by our resident guru d.o.g.

I just have one thing to add.
Being in manufacturing line previously, it is not so straightforward to make assumptions on number of Machines and the number of people tending to it. It very much depends on the company's manufacturing set-up.

Take a plastic injection machine for example, here is how it goes.
You need 2-3 persons to set up the tool/heat the raw material to rid of moisture etc.
These guys will be free once machine is set up. Sometimes these folks double up as tool maintenance technicians.

Once up and running, you may need one or sometimes no one tending to it. Depends on
how automated is the set-up.

At this stage, one person maybe at the machine to do some rectification on small defects.(i.e. cutting the edges clean)

At regular intervals, 1 Q.A. guy will do a sampling check for manufacturing defects.

Batch is completed, sent to the next process/stage of manufacturing....

[sgmystique]

QINGMEI’S FIRST QUARTER PROFIT UP 12.7% TO RMB80.5 MILLION

- Revenue rises 20.1% to RMB364.9 million
- Gross profit margin improves 1.5 percentage points to 30.5% on cost savings due to economies of scale
- Issue price fixed at S$0.135 for shareholders who have elected to receive part or all of the FY2011 Dividend in the form of new Shares under the Scrip Dividend Scheme; dividend payout of approximately RMB82.7 million in FY2011

Well, well, well...another very good set of numbers at first glance...but the stock price seems to be plumbing new depths...something just does not make sense here...