Belzberg Technologies Inc. (TSX:BLZ)



Google Finance


Dec. 20, 2007


Summery
This company's business is simple: Stock Brokerage. 
Analysis
1. Current market price around $5. Real book value is (32.5m-1.6m-0.5m-1.6m-0.3m) /14.8 = $1.93. Almost all in cash which is good.

2. Last years report earning is 61 cents pre share. Some adjustment:
Add currency lose 1.4m*69% = 1m.
Add stock buy back plan cost 0.2m. (BTW: it didn't happen, but the cost is there. This year end, they again passed 5% buy back plan. I think they will do it this time since the stock price is cheap now.)
Minus tax benefice 1.8m+0.8m=2.6m.
So roughly should deduct 1.4m from their reported income. Which reduce earning around 10 cents pre share. Actual earning will be 51 cents pre share.


3. This year's estimated earning pre share will be around 50 cents too.


4. Compensation and related benefits cost:
    2005    2006      2007
    8.1m    9.0m     10.9m(estimated)
This increased a lot and took too much of their revenue. This might caused by their acquire of Nandra Group in Sept. 2006. You can see their first 3 quarters of 2006 this number is roughly 2.1m each quarter. Started in 4th quarter of 2006, This number is around 2.7m. That's why 2006 increase 0.8m from 2005. 2007 increased 2.8m from 2005.


5. The admin expense will increase 0.8m in 2007 compared to 2006.


6. Because of 4 and 5, although their revenue in 2007 will increase more than 5m than 2006. Their income stay the same.


7. They issued quite a lot options(1.8m shares outstanding at 2006 year end. That is 13% of their total stock). The good news is at Sept. 30, 2007. There is only 0.6m outstanding with average price 4.39. (1.2m forfeited or expired). Thanks for the low stock price I think.


8. Their growth is depending on their new clearing service. I don't understand it very well. Just assume it doesn't decrease their revenue.


9. All above is not really bothering me. Only one thing left, which is also the most important, their revenue is largely depending on the number of trading orders they exercised. Obviously, they will do better in a bull market year since there will be more trading than bad year's. So I should estimate a average trading volume for them. However, from 2002 to 2007, they experienced a quite good growth. I don't know whether this is got by more new customers or just by increased volume from existing customer, or even both. So it is hard to estimate a average volume for them.


Conclusion:
If I take .50 cents as their average earning. At $4.50, I will buy it. However, due to the reason above, I can't do it before I figure out how bull market year and bear market year will affect their business. So this is not really a conclusion :)


Update:
Mar. 30, 2008

2007 Q4 Data


1. Recent market price around $4.5. Real book value (34.6m-1.0m-0.3m-1.6m-0.2m)/15.0 = $2.10. Almost all in cash.


2. 2007 net income 0.48 pre share.


3. Compensation and related benefits cost in 2007 is 10.8m.


4. The admin expense increased 0.95m in 2007 compared to 2006.


5. Revenue of Q4 2007 revenue decreased by 1.2m compared to Q4 2006. Revenue in 2007 increased by 3.3m compared to 2006.


6. Telecom and data fee in Q4 2007 increased by 0.1m. In 2007 increased by 1.0m.


Analysis:


1. In Q4 2007, its revenue decreased by 12% compared to Q4 2006. Which is obviously caused by current market downward. So the market condition did affect their revenue.


2. Cost of exchange, clearing and brokerage fee as percentage of revenue in Q4 2007 is 25% compared to 34% in Q4 2006. This is quite strange since Q4 2007 revenue actually decreased. In 2007 this number is 27.6% compared to 31% in 2006. This might be related to their new clearing service which could reduce cost. Any way, this is a very good sign.


3. The telecom and data cost now is around 6m pre year. This fee are pretty fixed and will not change when revenue grows or drop.


4. They added three new lines: (1) Full Service Self Clearing. (2) DeepBook(TM) for Options Traders. (3)BestX(TM) smart order routing and algorithmic trading. The first one is to attract new customers ans they said it did, the second and third are technique improvements.


5. In Jan. 2008, the company create an new RIN(Renewable Identification Number) online exchange market. The RIN is a credit which is attached to each gallon of renewable fuel produced in US like gas produced from corn or soy bean. Any gasoline sold in US if it doesn't use some percentage of renewable fuel(7.76% in 2008), it must attach the same among of RIN. The exchange charge 5% of each transaction from both seller and buyer. Assume the price for each gallon is 0.05. By 2022 this total market could be 36Bx0.05 = 1.8B. The RIN market itself is very interesting as well. However, this new line will not increase their income notably at least in this year since its trading volume is very low.

6. At previous analysis, I thought the market situation will affect their revenue. It is actually not too hard to check it out by simply go to yahoo finance and check S&P 500 and Nasdaq index volume history. I found out that for the past 5 years, Nasdaq volume stay as pretty as the same as 2B daily while S&P 500 volume doubled from 2B to close to 4B daily. So their revenue growth did somehow come from good market conditions in recent years.

7. However, although from last quarter to now the volume decreased by some, but I found historically the trading volume is not related to market conditions that much. They just keep increasing. Which means the current volume is most likely will keep fluctuating towards upwards than downwards. Which means their revenue will not suffer that much from current market conditions.


Conclusion:
Although the question in previous analysis is answered, however, the socket does not fit my value investment rules.
1. The book value is not close to price. 2. Two of Recent five years income are negative. I probably should create
rules for this kind of stock as well.
(Mar. 11, 2009: It is bad to create new rules because it breaks THE RULES.)


Update:
May. 12, 2008



Some analysis on Q1 2008 data:


1. Recent market price around $4.6. Real book value (35.9m-1.0m-0.4m-1.6m-0.2m)/14.9m = $2.19. Almost all in cash.
2. Q1 Earning 0.07 pre share. 0.11 pre share last year.

3. Compensation and related benefits cost 2.8m. Pretty the same as last year.


4. Admin expense: 1.1m. Pretty the same last year same quarter.

5. Transaction revenue decreased by 1.3m while exchange, clearing and brokerage fees saved 0.9m compared to last year.


6. Hired 7 new sales person in Q1.

Analysis:
1. Cost of exchange, clearing and brokerage fee as percentage of revenue for this quarter is 24% compare to 29% last year. The same thing as I noticed in previous quarter. They seems manage to decreased their cost of exchange fees. When their revenue getting better, this will save them quite a lot.


2. Compensation 2.8m. It has 100 employees. Aver $112,000/year for each person. They hired 7 new people in Q1. It is good this number hasn't increased.


3. Their revenue decreased compare to last year. Close to 2006 to 2005 level. But I didn't found volume of Nasdaq or S&P 500 decreased this year. I found this is really the most important thing about their business.



Aug. 15, 2008
Q2 2008 data:

1. Recent market price around $3.88. Real book value (35.5m-1.0m-0.5-1.6m-0.2m)/14.9m = $2.16. Cash 24.3m

2. Q2 Earning -0.03 pre share compare to 0.11 pre share last year.

3. Compensation 2.95m.

4. Admin expense 1.4m.

5. Revenue declined 19% from 11.1m to 9.0m compare to same quarter last year.

6. Lost $800,000 on a trading error. Will be write off at Q3 2008 which will be 0.03 pre share.

Analysis:
1. One big customer of floor brokerage cease trading this quarter count for 1.0m in revenue decrease.

Nov. 24, 2008
Q3 2008 data:

Bought some at $2.05. Recent price around $2.10.

1. Book value around $2.14. Cash 18.5m. Cash $1.24 pre share.


2. Q3 earning 0.0 pre share excluding 0.8m trading error.


3. Compensation 3.16m.

4. Admin expense 1.25m.


5. Revenue increased by 14%, however, gross margin is down.


6. Released a news after Q3 data that in Oct. it get extra 0.8m on options trading and good sales revenue at that month. Will be reflected in Q4 data.

Mar. 10, 2009
Q4 2008 data
Current price $1.87


1. Hired new CEO and president.


2. Restructuring charge $1.1m (paid to former CEO). Unusual gain on foreign exchange 0.8m reported on Nov. 2008


3. The actual earning of Q4 is -0.7m before tax. 0 after tax.


4. Revenue $7.6m($12.2m including pass through ) compare to 6.9(9.0 including pass through) same quarter last year.


5. Cost of revenue ( 0.8(5.4)+2.1) = 2.9m compare to (0.1(2.2)+1.5)=1.6m. Gross profit $4.7m compare to $5.3m last year.


6. Admin $1.5m compare to $1.3m same quarter last year.7. Compensation $3.6m. Over $1.0m above previous year. May related to hire new CEO and president. It is pretty unacceptable given their current income level. I should be alerted when their Q2 compensation increased to 3.0m and Q3 increased to 3.2m.


Other thoughts:
1. Although we should not take one quarter of data as big deal. The new management team might need more time to get things better. But until now they show an very weak management. The market is bad now but they have not intention to cut cost and there is no sigh their cost will be reduced instead of keep increasing.


2. I brook two rules to buy this stock: "Two of Recent five years income are negative." & "They never paid any dividend". Obviously breaking rules is bad.


According to above, It is a mistake to buy this stock. Fortunately I bought it below book value and they have a lot of cash. Hopefully I can sell it without big lose.




Feb. 23, 2011
Recent price: $0.50


A company called Frontline Technologies Corp. will merge with the company. BLZ will pay FrontLine $1m in cash and issue 7,881,826 new shares to Frontline. Exchange for around $2.75m in working capital and $4m annual sales, $0.2m annual income. Existing BLZ business will be closed and proceed will be distributed to existing share holders only.

1.Current share outstanding: 14.638m, New shares after merge:  22.52m. With a new working capital around $2.75m  and  $0.2m annual income, it worths at least $0.10c or more. Actually Frontline must expect it should worth ($2.75m-$1)/7.882m = $0.22 per share to make a even.

2. Based Q3 2010 data, BLZ has $13.5m in equity.( $12m in working capital. $1.25m in capital assets. $14.5m in cash. $3m in liability). Minus $1m cash will be paid to Frontline. It is $12.5m on paper for distribution at Q3 2010.

3. Four big share holders is now on board.  Total 5.3m shares, 36%.


After 2 quarters of lost, if some how the $12.5m eventually can be liquidated over $6m. Then there are some profit there. 

Mar. 03, 2011
Bought more at $0.39

The asset must be liquidated for $4.25m and after merge stock price must be $0.10 for me to make an even.


Mar. 14, 2011
Got the agreement of Frontline deal and agreement of ref customers to EBS.

1. Frontline current working capital is likely to close to $0(range from -$300k to $300k). Which means the $2.75m will be deducted from BLZ current cash. This amount will be increased or reduced by the final number of Frontline's current working capital.  

2. The refer of customer to EBS will be started before Apr. 1st, 2011. Within one year, the operation will be transferred to EBS gradually and EBS will pay BLZ for the operation cost during the period. 

3. The deal with Frontline seems was directed  by Cameron MacDonald, the biggest share holder of BLZ (2.1m shares). He is the CEO of Goodwood fund. It is value fund which returned quite good since inception(1996). He is also CEO of Westaim corp(TSE:WED) which Goodwood holds 20% shares of it. The Westaim is very similar to BLZ in nature.

Recalculation:


1.BLZ will give up $3.75m(0.26 per share)cash in this deal. Assuming the deal is fair, then it is reasonable to think the new company will worth $0.26 per share. Otherwise, BLZ could just distribute the cash to share holder in stead of going though this whole merger process.  


2. In total, BLZ will pay $1m + $2.75m*35% in exchange for 65% of the equity of Frontline. If the deal is fair based on numbers, Frontline could have ($1m+$2.75*35%)/65% = $3m in equity before merger. However, I don't expect that they would have that amount of long term assets. It it is more reasonable to range from $1 to $2m, then future book value will be around ($2.75+$1)/22.52=$0.17 to ($2.75+$2)/22.52 = $0.21.  


3. The existing long term asset of BLZ is $1.3m, plus 150k tax credit.  The U.S subsidiary are very likely be liquidated while the Canadian office remains unknown. Let's assume this could be added $0.05 to distribution and future stock value in total. 


4. The referral of customers to EBS might return quite a bit for BLZ, also it will happen before Apr. 1st, which is much earlier than expected. let's take Saj's assumption the is will nets out the cost of closing cost to zero. 


5. The total amount of working capital available for distribution at Sept. 2010 will be ($12-$1-$2.75)= $8.25m.  Assuming it had lost 4m in past 2 quarters, now it has $4.25 in cash for distribution which is $0.29 per share.


1+3+5=$0.26+$0.05+$0.29=$0.60 per share. The price I bought $0.39 is a 35% discount.  Ideally it should be no more than $0.36 which is 40% discount.


Comments:
I had made a mistake again in this stock which is too rush to buy. The two reports was released just 2 days after I bought the stock. Should have been more patient and wait its release. 


Luckily the price I bought wasn't that bad. For Saj's opinion, he is right on the $2.75m working capital while missed the value of the new shares and the long term assets. Personally I trust Cameron MacDonald wouldn't let us down. Will continue to hold it.


http://www.barelkarsan.com/2011/03/belzberg-uncertain-payout.html

Mar. 22. 2011
Q4 2010 release and Annual Information Circular.
(1)At Dec. 2010, BLZ had working capital $10.3m. $Assets around 1.1m. 

(2)I was wrong thought this is a reverse merger. Actually BLZ will have some operation with new company. The new company had a $6.6m pro forma revenue in first 3Q of 2010. 

(3)At Dec. 31, 2010, Frontline had equity around $330k. Working capital (200k). 

(4)Frontline currently has 34 employees.

(5)Astaraki the owner of Frontline declared bankruptcy at 2003. It said was caused of his parents debt.


(6)The distribution will be decided only after one year of deal close. 


(7)The current CEO will leave with $300k compensation. 


Recalculation:


1. BLZ will give up around $3.55m. $1m to Astaraki and leave $2.55 to new company. Around $0.24 per share.


2. After merge. It will have around $2.8m in equity and working capital around $2.35m.  


3. Some of the $1.1m asset from BLZ will be liquidated and some will go to Frontline.  Assume half will go to Frontline, half will be liquidated. 


4. Available fund for distribution at Dec. 2010 was $10.3-$3.55+$0.5=$7.25m. Assuming $2m more loses. Now it has $5.25 for distribution which is $0.35 per share.


Around $0.60 in value. Not much change from previous calculation. 


Thoughts 
1. Frontline currently only have $300k in equity.  It is much less than I thought.  It means BLZ have trade around $1.8m just for $300k*65%=$200k in return. They might just liked the new business.


2. The new business will have $8m to $9m in revenue which should be able to support an $0.20 to $0.30 in share price. However, it must be remain profitable to get that. 


3. Bad part is now I feel the distribute is very uncertain. Because these guys seems care more about the new company than the U.S liquidation. Plus the new CEO will get nothing from the distribution, he would care nothing about this matter.     



Disclosure: Author has a long position in BLZ

Dover Saddlery, Inc. (DOVR)


Dec. 13, 2007 
Summery
When I first find this company, I am quite excited because I thought I have found the kind of company I am keep looking for. This company is in a unique market. Targeting luxury customers. No strong competitors. Well grows strategies. However, after reading their financial report carefully, I found that their finance result is not very good. That's why they are very cheap now. Generally I like this company and would invest at a proper price. 
Analysis.
1. The historic average pre share earning is 25 cents. But this doesn't mean a lot since they are on the market for only 2 years.
2. Book value is only $0.87 pre share.
3. Their current year earning is pretty bad. The first three quarter earning together is -3 cents pre share. A important reason for this is they are involved in a litigation settlement which cost them 0.7m in the first quarter. That money means 14c pre share.
4. It is all about Selling G&A. When I exam this company, I found there is only one thing is really bothering me. In the last three quarter, their selling, G&A expense is lot of higher than in 2006. So I need to figure it out. Let's show the number first:
             Q1        Q2        Q3         Q4
2006     5.4m     5.5m      5.7m     7.3m
2007     7.1m     6.6m      6.2m        ?
You can see from Q4 2006, their Selling G&A expense is in a big hike from 5.7m to 7.3m. I found out the reason is they had opened a new store at Q3 2006. There are a lot expense related to this new store are gradually down during the last 4 quarter.  if we take Q3 2007 number as average number. Then this year, we can deduction ( 7.1-6.2)=0.9 for Q1. (6.6-6.2)=0.4 for Q2. So this well contribute 26 cents to their pre year earning.
They opened a new store at Q3 2007, which I can expect there will be another hike in their Q4 2007 expense. But it will not bother me.
5. I estimate their Q4 2007 pre share earning is at lest 10 cents base on historic data.
Risk
1. Their opening new store strategies doesn't go well. This is not very likely based on their last years new store and this years revenue increase.
Conclusion:
Added all 3, 4, 5 number, their current year real earning would be (-3+14+26+10) = 47 cents. This didn't count their future grows by opening new stores. Based on this calc. I would like to pay up to 0.47x15x60% = $4.23 for this stock. This is much higher than current market price. I will be hurry.
Updated: 
Jan. 7 2008
It turns out that I had made a mistake on this stock(DOVR). Luckily I was able to sell my holding with a speculating 7.7% gains. The reason is below:
1. Its book value is too low and they are working by very high debt level. According to Buffet theory, the price paid should not exceed 120% of a stock's book value. I think I'd better stick to it.
2. Its history on the market is too short( 2 years +) , they have deficit years and don't have any dividend.
The stocking might be a good one. But for sure it doesn't fit my list. I liked their business too much. That's why I neglected all these important stuff. Another reason is I haven't formed these rules when I first assessed it.
There is the same problem in VG(Vonage). However, VG belong to another category: special opportunity according the book. So these rules is not important to it. I think I will still hold VG. But probably will not buy the same stuff any more. 
Updated:
Mar. 25, 2008
Analysis of Q4 2007 data:

Facts:

1. Per share earning is 0.18. Better than my 0.10 estimation. New store sales contributed to it.

2. G&A of Q4 2007 is 7.3m, the same as 2006. As I have expected, it is much higher than previous quarter.

3. Share outstanding 5.2m.

4. Book value around $1.

Analysis:

1. Take a 6.7m as it is future average G&A number per quarter. Q4 2007 should add (7.3-6.7)/5.2 = 0.12 pre share.

2. If take the new 0.18 Q4 earning and add the 0.12 quarter G&A data. It is real earning in 2007 should be -0.3+0.14+0.26+0.18+0.12=0.67.

3. However, In the near future, they probably hard to achieve this earning since they are keep opening new stores. There will be more expense there.

4. The retail store strategy seems works because usually Q4 is their weak quarter, but it achieve 0.18 pre share earning in Q4 2007 with bad weather last year. ( Aug. 26th: This is wrong, actually Q4 is their strongest quarter because it is the holiday season. Q2 is second strong quarter. Q1 and Q3 are weak quarters).

Conclusion: I think it worthies much more than their current market price( around $4). However, I can't buy it because its book value is too low.
Updated: 
Aug. 25, 2008

Q1 and Q2 2008 data, recent price $2.80.

1. G&A data:
             Q1        Q2        Q3         Q4
2006     5.4m     5.5m      5.7m     7.3m
2007     7.1m     6.6m      6.2m     7.3m
2008     6.6m     6.3m

2. Book value $0.94

3. Refinanced 5m notes at 14% interest ant end of 2007. Used $8m of $18m line of credit.

4. Income Q1:  -0.07 pre share. Q2: 0.05 pre share. Revenue declined a little bit compared to last year. Catalog sales drop while retail sales up.


Updated
Nov. 12, 2008

Q3 2008 data, recent price $1.63.  Market cap now only 8.5m. While annual sales around 70-80m.

1. G&A data:
            Q1        Q2        Q3         Q4             Total
2006     5.4m     5.5m      5.7m     7.3m          24m
2007     7.1m     6.6m      6.2m     7.3m          27.3m
2008     6.6m     6.3m      6.3m

2. Opened a new store at Georgia at Oct. This might increase G&A expense for Q4.

3. Book value $0.97

4. Total debt 14.3m. Last quarter 12.8m. It is flat compare to last year. Historically, Q3 debt should be higher for seasonal reason.

5. Income 0.03 pre share. Revenue flat. But same store sale decreased by 5.8 percent.

6. Inventory 17.7m. Close to sales of one quarter.


Analysis:

About previous analysis on G&A data, actually it was wrong. For seasonal reason, their Q4 G&A data should always much higher. It is more accurate on a yearly bases. 2008 seems they can reduce the G&A expense by 1m compare to last year. Which seems caused by more store opening at 2007. However, the revenue will decline in 2008 as well.

It is closer to its book value. Glad I sold the stock. According to conference call, they are very optimistic about their business. I think they are doing OK. Wait for the price get close to book value.

Mar. 30, 2009
Q4 and full 2008 data. Current price $2.15.

1. G&A data:
            Q1        Q2        Q3         Q4             Total
2006     5.4m     5.5m      5.7m     7.3m          24m
2007     7.1m     6.6m      6.2m     7.3m          27.3m
2008     6.6m     6.3m      6.3m     7.1m          26.3m

2. Opened two new stores at Oct. 2008 and Feb. 2009. According to conference call, no more new stores in 2009.

3. Book value $1.08. Impaired $14.3m goodwill.

4. Totol debt 13.2m. compared to 11m last year.

5. Income 0.07 pre share. Revenue decreased by 6.8% to 21.4m.

6. Inventory 17.3m.

Advanta Corp.(ADVNA ADVNB)


Dec. 11, 2007

Summery

This company has only one pretty simple business: Credit Card. However, it is not very simple when around 4/5 of their credit card receivable were packaged and sold to other party. I spent quite a lot of time try to understand how this ABS is issued. Until now I can not understand it thoroughly. But I also found what I understood is enough to draw a conclusion.

This company is well running for many years. I found there is only one thing that should be figure out: how charge off rate will affect their earning?  The charge off is lost of credit card receivables when card holder failed to pay it. This rate is vary from year to year. A slight change in this rate will affect their earning significantly.

Facts:
1.Their average charge off rate is (2002 to 2007) (7.92%+7.42%+6.38%+5.37%+3.17%+3.19%)/6=5.58%

2. In their 2007 Q3 report, it mentioned that if current lost rate increased by 10%, it will decrease their net income by $5.91m.

3. The estimate current year earning is $1.90 pre share. Shares outstanding 41.1m.

4. The average earning pre share is $1.30 for the last 6 years(2002 to 2007).

5. Book value (591m -(218mx73%))/41.1 = $10.51 ( I charge off 73% on their subprime investment)
Analysis:

You can see the main reason the company has been doing quite well since 2002 is the charge off rate keep decreasing. Last year(2006) their lost rate is in the lowest. This year it keep increasing.  And it will be going higher next year as a result of current credit crisis.

Let's take the average charge off rate as their future lost rate:

Then it means their lost will increase by ( 5.58-3.19)/3.19 = 75%.
Their future lost will be 5.91x(75%/10%)=44.3m.
Pre share lost will be 44.3/41.1=$1.08.
Future pre share earning might be $1.90-$1.08=$0.82.

This number is much lower than their average pre share earning $1.30. The reason for this is during the past several years, their get quite a lot income from two discontinued operations and other stuff. I don't even bother to calc those stuff.

Risk:

1. Charge off rate increase, already included in above. In a long run, the rate should be medium number, not high as 2002 or low as 2006.

2. They have 4/5 of their receivables are packaged as ABS. In those ABS structure, they may have problem to issue the B, C, D class(lower credit). Which will make themselves to bear more charge off loses. This is very possible as current subprime crisis is getting worse. However, in a long run, it should not be a problem since the underlying product is good.

3. Fed rate increase or LIBOR rate increase. They benefit from rate cut while lose from rate increase. In a long run, it is not a problem.

Conclusion:
At price (0.82x15)x60% = 7.38. It s a bargain. Since they have two share class A & class B. Their weight is 1:2. Their price difference is around $0.78. So the bargain price for A share is 7.38-(0.78*2/3)=$6.86. B share is 7.38+(0.78*1/3)=$7.64. They are quite lower than their current market price. Let's wait to next year.

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Updated:
Mar. 04, 2008


The new Q4 2007 data has been released for a while. Finally got some time to take a look.

Facts:
1. Q4 earning is $0.17 per share. Including $0.39 charge off based on credit trend. $0.11 charge on litigation and $0.05 gain on selling master card share. Whole year earning $1.61 pre share.

2. New charge off rate for 2006 is 3.48%. New charge off rate for 2007 is 3.79%

3. By Q4 2007, when charge off rate increase by 10%. Lost will be 6.51m. For 2006, when charge off rate increase by 10%. Lost will be 4.23m.

4. Share outstanding 42.5m.

Analysis:
1. If take off all the extra items. Earning in 2007 is 1.61+0.39+0.11-0.05 = $2.06. It is 0.16 higher than previous estimation.

2. Q4 charge of rate 4.13%, Q3 charge of rate 3.87%. Based on my calc, the lost in Q4 caused by charge off should be 5.91x((4.13%-3.87%)/3.87/10%)/4 = 0.99 million. Around 2.5c pre share.

3. In my previous analysis, charge off rates of 2006 and 2007 are 3.17% and 3.19%.  These changes in rate will result one time write off of  ((3.48-3.17)/3.17/10%)x4.23 + ((3.79-3.19)/3.19)/10%)x6.51 = 4.14 +12.24 = 16.4m.

4. Now take a look at the 0.39 pre share charge off in Q4 2007. It is 0.39x42.5m = 16.6m. It is close to the 16.4m calculated in 3. Basically that is what this charge off come from.

5. New average charge off rate is (2002 to 2007):  (7.92%+7.42%+6.38%+5.37%+3.48%+3.79%)/6=5.73%

6. Take off the 12.24m charge off in 3. Real earning in 2007 is 2.06-(12.24/42.5) = $1.77.

7. Estimate future earning based on average charge off:  1.77 - (((5.73-3.79)/3.79/10%)x6.51)/42.5 = $0.99. It is 0.17 cents higher than previous estimation. If I use 1.90 instead of 2.06 in 6. I will get $0.83. Almost the same as previous estimation.

8. Need to figure out why there is 0.16 cents difference in 1.


Conclusion:

At the end of the analysis, I was trying to figure out how this 0.16 cents difference was created since they are not getting better. Then I realize it is a mistake to invest this company. Not because I found something bad about it. But I found I am not really understand their finance statement. I also found I might better keep away from any financial stock for now because my accounting knowledge is not enough for understanding them.

Will sell it in a appropriate time.

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Updated:
Apr. 08, 2008

After listen to its Q4 conference call and read its annual report more carefully. More information about the mis 0.16 cents:

Analysis:

1. Extra-odinary items in Q4:
    (1) Pretax charge of 10m( 0.14/share) for owned receivables.
    (2) Pretax charge of 17.2m (0.25/share) for securitized receivables.
    (3) Pretax charge of 7.8m(0.11/share) for litigation related with Visa.
    (4) Pretax gain of 3.3m( 0.05/share) for selling master card share.

2. Compare of Q4 2007 and Q3 2007 income
                                                   Q4           Q3
   Interest Income                         49.8m      50.0m
   Interchange Income                   69.3m      62.7m
   Securitization Income                10.4m      22.4m  (2)
   Servicing revenues                     25.3m      24.2m
   Other revenues                            9.9m       8.6m   (4)
   -----------------------------------------------------------------------
   Interest Expense                       22.2m      24.4m
   Provision for credit losses           21.6m     14.7m              (1)           
   Operating expenses                   81.0m     68.1m (72.3m)  (3)

   (1) Pretax charge of 10m( 0.14/share) for owned receivables should be included in "Provision for credit losses". However, the number seems not match.
   (2) Pretax charge of 17.2m (0.25/share) for securitized receivables is included in "Securitization Income". However, the number seems not match, they didn't provide detail information in it.
   (3) Pretax charge of 7.8m(0.11/share) for litigation related with Visa is included in "Operating expenses". In Q3 it is 4.2m, which shown in the adjusted 72.3m number.
   (4) Pretax gain of 3.3m( 0.05/share) for selling master card share is included in "Other revenues". In previous quarter it is 2.9m.


3. But it is not that hard to understand the where 0.16 come from now.  Increase in interchange income 6.6m plus 2.2m saved in interest expense should contributed to it. Seems just a seasonal stuff since there are more transaction in Q4, not really important.


Conclusion:

Its financial report is hard to  understand. Although I get better idea how their charge off is calculated, but still quite confusing. Need more time to figure it out.



Updated:
May. 02, 2008

Q1 2008 data has been released.

1. Charge off rate increased to 6.43%.  Real earning 16c pre share.

2. A quick calc using previous quarter's formula estimated yearly earning should be 1.77 - (((6.43-3.79)/3.79/10%)x6.51)/42.5 = $0.70. Not much different from 16c x 4.

3. The charge off rate has increased quite a lot. It is already above the average number 5.73%. I didn't expect it raising so fast. Neither did they.

4. From their conference, Florida and California's lost rate are 1.3 times of other area. Business of these two states are account 24% of total business. This is a important sign that Advanta's business is more connected to housing market than I previous expected.

5. If all other area's lost rate increase to the same lever as in these two states. Their charge off would be 6.43% + (6.43%x0.3x76%) = 7.9% which is the level of 2002. At that level. They will start to loss money. But I am afraid it might getting worse. At 9% lost rate, they will lost around 35c pre share yearly. At 10% lost rate, they will lose 75c pre  share yearly.

6. Taking their high book value, those number are not too bad even it did happen like that. Unless this crisis getting very bad and last for long time.

Conclusion:

Somehow I have sold this stock with 16% profit. For holding just four months, I can't complain at all.
1. At my current knowledge level, finance stock is not very suitable target since they are not easy to understand throughly
2. The reward of this stock potentially might be very high. However, it has much bigger risk than my other holdings. I am still quite new to the market. Better keep myself safe.


Nov. 23rd, 2009
I can't tell how lucky I was after the company filed for bankruptcy earlier this month and was delisted from NASDAQ now.  There was a time that the company was selling at 20c and I was tempted to buy it back. Fortunately I didn't do it because I missed the right price point I want. Over all, it is so important that I should not get into a business that I can't understand its operation.





Deswell Industries, Inc. (NASD:DSWL)



SEC Filing link

   (Year end Mar. 31)  
   Nov. 28. 2007


Summary:
The company suffers a lot of decrease in earnings in the recent two quarters. They had a quite good earning history. It says in their quarterly report that the reason is from three cost increase: Resin price, Labor, and RMB appreciation. I am trying to figure out each of them counted for how much of the losses. In my opinion, the resin cost increase should not be counted as a permanent factor to their business. They already start a price increase based on it.
Q2 2008 (End Sept.30, 2007) revenue decrease analysis:
Assumptions:
1. Based on their employee number of 4000 in plastic product and 2400 in electronic product. Assume they have 3800 workers and 2200 worker in both segment respectively.
2. Their worker's average cost is CNY1500 a month.
3. Resin price increased by 15% in 2007 compared to 2006. (I searched on some website and estimate this number. It turns out working very well in the following calc).

Three months compare to Q2 2007.
1. Total cost increased 2.9m.
2. Labor cost increased 3800x(1500/7.5)x3x17.2% + 2200x(1500/7.5)x3x22.7% = 0.7m
3. RMB appreciation increased cost around 38.4x2.7% = 1m.
4. Resin cost increased around 38.4x20%x15% = 1.15m.

Six months compare to Q2 and Q1 2007
1. Total cost increased 6.3m.
2. Labor cost increased 3800x(1500/7.5)x6x15.9% + 2200x(1500/7.5)x6x30.2% = 1.5m
3. RMB appreciation increased cost around 76.9x3.1% = 2.4m.
4. Resin cost increased 76.9x20%x15% = 2.3m
As you can see, the 2+3+4 is so close to 1. Except for my 3 assumptions. All the other numbers are from their reports or calc from their reports.




Fact:
1. The last 3 quarters real earning is 30 cents. Last 4 quarters real earning is 54 cents. If this trend continues. Their earning will be around 40 cents per year. But they will increase their price for sure. This should not be used to calc their intrinsic value.
2. Book value now $7.47 per share. If take a 50% discount on their plant $60mx50% = 30m, It still worthies $5.57.
3. They don't have any debt.
4. They pay quite a lot of dividends and has a long dividend history.
5. Compared to 2005. They had a big customer(vTech) didn't have a contract with them in 2006. Another (Epson) didn't have a contract with them in 2007. This might count 10%-15% of their revenue. That is why they performed better in 2005. In 2007, they added Peavey(10% of revenue). But it is in the electronic segment which has the low margin(10%).

Risk:
1. They just had a new CEO try to put more weight on electronic segment(now 55% and still increasing) which has a low profit margin(10%). They also put a lot of weight on R&D their own audio device product. Their growth is depending on the performance of their electronic segment and their audio device. If it doesn't go well, their might suffer big losses. According to their past, I believe their chance to do well is higher than go worse.
2. The RMB continue to appreciate against the $US and labor cost continue to increase. This is almost for sure but will be more modest and will not affect them that much as in current year. It will definitely affect their margin in the future.
3. Resin price continue to fluctuate. They will gain or suffer lost. But in a long run, this should not be counted as a permanent factor. Since every factory will has the same problem and eventually the cost of resin will be reflected in the final product.
4. Their price increase drive their customer away. It is possible if they increase their price too much. A modest raise based on resin cost should be fine.
5. Their big customers hold their contract for some years. In a single year, it has a bad effect. In a multiple year base, it should be fine.

Conclusions:
1. Assuming their resin cost increase should be covered by price increase in product. This will give them a 29c/year more in revenue. Plus current 40c/year revenue. They will have a 69c/year revenue.
2. Due to the appreciation of RMB and increased labor cost( this should not be added to final product price). It is hard for them to get the average 80c/year earning in the past.
3. This give them a intrinsic value of 0.69x15 = $10.35. At a price of $6.21 (40% off), it is a bargain.

Updated:

Feb. 29, 2008

Q3 2008(End at Dec. 31, 2007) data has been released. Let's see whether previous formula still works.

Three months compare to Q3 2007.

1. Total cost has not increased.

2. Labor cost increased 3800x(1500/7.2)x3x12.9% + 2200x(1500/7.2)x3x22.6% = 0.6m

3. RMB appreciation increased cost around 35.4x2.7% = 1m.

4. Resin cost is not mentioned..

Nine months ending compare to Q3 2007

1. Total cost increased 6.3m.
2. Labor cost increased 3800x(1500/7.2)x9x15% + 2200x(1500/7.2)x9x25.6% = 2.1m
3. RMB appreciation increased cost around 112.3x3.2% = 3.6m.
4. Resin cost is not mentioned. use previous 6 months data: 2.3m

Facts:

1. Real earning is 2.7m. Which is 17c/share.

2. Q3 Revenue: 35.4m.

3. Share outstanding: 15.8m

4. SG&A increased by 0.3m.

5. Dividend 17c.


You can see the cost calc is not working anymore. In Q3, The Labor and RMB cost count for 1.6m. But the overall cost has not increased. The night months cost also has a 1.6m unbalanced.

If takes off the 1.6m from its 2.7m Q3 earning. Add the 0.3 SG&A cost increase. It would have 1.4m. Which is 9c per month. Added previous 3 quarter's 30 cents earning. It come out as 39c. Not too far from my previous estimate 40c pre year.

So revenue improved by 1.6-0.3=1.3m this quarter. It is close to my calc of how much they could get from price increase only. However, according to the report, it is a combine of closing low margin productions, reducing manufacturing costs, changing in customer and product mix, price increase on some production. This means price increase is not ended yet. It mentioned in the report that there will be a price increase in the coming quarter.

Conclusions:
They are doing better than I have expected. Continue to hold.

Mar. 03, 2009
Q3 2009 data
Recent price $1.6. For the past several quarters:

Q4 2008 income: 0.12 per share. dividend 0.12 per share.
Q1 2009 income: 0.08 per share. dividend 0.08 per share.
Q2 2009 income: (0.11) per share. dividend 0.00 per share.
Q3 2009 income: 0.06 per share. dividend 0.04 per share.

June 2010
Q3 2010 data
Recent price around $3.75. This is the longest I have hold(since 2007). Redo the check list since previous study were pretty useless.

Check list:
1.Major Business.
Funded in 1987 by Richard Lau, Chin Pang Li, Chi Wai Leung. IPO at 1996. 
Plastic injection molding: 50% -- 50%  JetCrown Industrial.
Electric and Audio device: 50% --  50%  Kwan Hong Electronics.

Major customers include:
N&J Company Limited (Nintendo (NTDOY.PK)), 
Digidesign, Inc. (pro audio), 
VTech Telecommunications Limited (wireless telephones), 
Inter-Tel Incorporated (communications solutions), 
Peavey Electronic Corp. (pro audio)

2.Price/Book ratio.
Book value 123m/16.2m = $7.59. 

3.Current ratio. Debt/Current Asset ratio. Debt maturity and interest. Inventory level.
Currently no debt. Cash per share $2.81. Current inventory 16m.

4.P/E ratio. Deficit check. Sales/Price ratio.
Average earning for past 5 years(2009-2005): (0.08+0.57+0.81+0.59+1.02)/5 = $0.61
Current Sales/Price: 90m/60m = 1.5.

5.Dividend history.
2005-2009:  $0.65, $0.63, $0.65, $0.61, $0.24.
Q1 - Q3 2010:  $0,   $0.10,   $0.

6.Free Cash flow
                      2005         2006         2007         2008       2009         Q1                Q2                 Q3
FCF :             $11.3m   $5.4m      ($0.5m)      $9m       $4.6m       $6.3m       $11.4m*        $0.7m
Dividend:     $10m       $8.4m      $12.4m      $9.5m    $3.8m       $0              $0*                  $0
*Q2 2010 FCF includes $5.5m from sale of Shenzhen building. Also around $1.6m dividend payment due is not recorded.
Since 2001 the company invest over $57m on DongGuang manufacture building. Phase 1,2,3 finished. Phase 4 is likely postponed currently. 

7.Grows related.
Revenue from 2005 to 2009: 125m, 115m, 136m, 143m, 131m. 
First 3Q of 2010 sales down 40% to 65m compare to 104m last year.

8.SG&A, R&D expense.
No big hike on SG&A number. In a down turn, management would take action to cut cost.


9.Management compensation. Options. 
2009: $1.9m for 7 executes (include 2 retired and 1 died). 
2008: $2.5m for 7 executes
2007: $3.2m for 7 executes
2006: $2.2m for 6 executes
2005: $3.5m for 7 executes (include one retired director)

10.Buy back, insider holding and trading info.
Richard Lau( Chairman, former CEO) 1.35 shares. 8.5%
Chin Pang Li(Admin for Plastic): 1.15m shares. 7.3%
Chi Wai Leung:(Engineering for Plastic): Resigned at Jan. 01, 2009. 
Shu Kwan Lee: (Previous CEO of electronics):  Retired at May. 01, 2009
Man Chi Tam: (Manufacturing for electronics) died at Feb 23, 2009.

11.Employee numbers. Revenue/Employee. Compensation/Employee.
not important 

12.Industry comparison.
Nam Tai Electronics (Ticker: NTE)

13.Major events.
Feb. 2007: Frankie Tse succeeded Richard Lau as new CEO.
Aug. 2008: Betty Lam replace Eliza Pang as new CFO.
Dec. 2008. Founder Chi Wai Leung resigned at age 53. Addressed personal reason to relocate oversea. He has 1.15m shares at that time. Reduced to below 5% at June 30, 2009
Feb. 23, 2009: Man Chi Tam died at age 59. He has 362k share at that time
April 29, 2009: Shu Kwan Lee retired at age 62. No share number is mentioned. At 2008 his share is around the same as Man Chi Tam.
Mar 31, 2010: Reduce quorum of meeting from 50% of total shares to 33.3%.

14.Others.
About land and building:
(1)ShenZhen 10.368k sqm factory building 
At 2000 bought at RMB 12M. (RMB1157/sqm)
At 2009 sold     at RMB 50m. (RMB4816/sqm).

(2) DongGuang HouJie 118k sqm land and building
At 2000 bought land for RMB16m. (RMB135/sqm). the usage is up to Dec. 31, 2049. 
Normal management fee is RMB1.15m/year. 
Building and hardware( in US$ and square feet )
             Construction  Machine  Furniture   factory  amenity  dormitory  office
Phase I:  $8.0m          $19.9m    $7.4m      466k     91k        116k 
Phase II: $7.3m          $4.7m      $2.1m      227k(2)              216k(2)
Phase III:$6.9m                         $1.6m      377k                  120k         133k
-------------------------------------------------------------------------------------------------------
Total       $22.2           $24.6m    $11.1m    1070k    91k       452k         133k  
Total building: 160k sqm 

(3) DongGuang ChangAn 22.4k sqm land and 38.5k sqm building.
At 2003 bought at HK$25m(HK$650/sqm), totally $4.2m recorded.  
Management fee is HK$270k/year.

At Dec.31, 2009, total recorded at cost:
Land and Building: $34.7m  (Total building = 160k+38.5k = 198.5k sqm. Cost: $174/sqm)
Machine: $40.6m  Furniture: $24.4m  Others:   $7m.
Total net book value recorded : $62m.

Given China't real estate appreciation, currently land and building might be liquidated at least 2 times of the cost. 

15.Concerns.
There is a good article in seekingalpha:

(1) The 3 executive officer vacancy still hasn't been refilled. Sales down 35% after they left. Don't know whether they are related.

(2) Dividend had been suspended for Q1 and Q3 2010.

(3) Royce & Associates, LLC keep decreasing position in DSWL.


July, 22, 2010
Q4 2010 data


Bought more at $3.56.


1. Sales down 39% compare to last year.

2. Lost around $1m for this quarter.

3. FCF for whole 2010 year is 12.9m ( including $5.5m cash from sale of Shenzhen building).

4. Dividend 5c.


Dec. 20, 2010
Q1 and Q2 2011 data

Recent price 3.45

1. 2010 all executive compensation $1.2m. Including compensation for Shu Kwan Lee who retired at May 01, 2009.
2. Sales in Q2 is up 15% to $24m, however, it is still much lower than 2008 ($32m) and 2007 ($38m) level.
3. Lost 0.24 per share for Q1 and Q2.
4. Cash decreased $8m compared to the 2010 year-end. However, despite working capital changes, it is just a little decrease. Actually, Q2 is positive.




Feb. 04, 2011
Bought some at $3.05.


Mar. 10, 2011
Q3 2011 data
Recent price $3.24
1. Q3 income excluding the $3.6m write down of assets, should be just made even.
2. Operating cash flow in Q3 is around $1m excluding working capital changes. First 3Q FCF excluding working capital changes made even.
3. Cut headcount from 4100 to 2900.

July 09, 2019
Price $2.72 Shares 16m. Cap: $43m.
2019 data(2019.03)

(1) The company has recovered for the last two years. Now made around $1 to $2m in operating plus $3m to $4m in investment gains.
(2) Current cash is around $38m. Book value is around $84m. Currently paying a 0.14/year dividend.
(3) CEO Richard Lau kept buying stocks. Now he owns over 50% of total shares.
(4) The company is still weak in terms of operating. Income relies heavily on investment income which is a little high for the last two years.