Right now I own some Jan 2000 30s, but am looking to get some more LEAPS, you think the Jan 1999 25s are a good buy? I think I may get some more 2000s, maybe the 25s or more 30s.
I am going to update the LOR FAQ, idt.net
today or tomorrow, I need to finish recovering from New Years and the Rose Bowl. But wait! It's Friday :(
Here are Readware's comments, all of them for the last week or so, sorry for any repeats:
===========================================
Subject: A Good Sign
Date: Thu, Dec 25, 1997 19:02 EST
From: Readware
Message-id: <19971226000201.TAA09600@ladder02.news.aol.com>
The Iridium Boeing launch at the Cape, scheduled for 4 February (the day before G*), has been moved up to 31 January. This is a good sign-- it means that the Boeing launch facilities have to be readied again for 5 February, and now have the time for that. The Iridium launch being originally just a day before, as was originally scheduled, did have me thinking about 5 February and if it could in fact be the G* launch window, given the shortness of
time-- one day.
The change is an increasingly surer sign that G* has 5 February as its launch window.
Subject: Re: ICO/READWARE
Date: Sat, Dec 27, 1997 13:36 EST
From: Readware
Message-id: <19971227183601.NAA09299@ladder01.news.aol.com>
ICO has regulatory advantages that G* and IRIDF do not have. So they will be able to clear FCC ITU licenses more quickly than the other two were able. Their technology is not anything causing concern, and I believe they will use TDMA for phone voice. Now that they are working with TRW I do not know if they will use a MEO-GEO system. TRW gave up its CDMA spectrum, so I do not see ICO using that. The date of their actual service depends on the
timeliness of successful launches. They plan on being in service by 2000. Do not forget, however, that Hughes already has a GEO telephony system planned in the Middle East-- the launch timeliness of the GEOs involved determines the actual service date, although I believe that they plan initation in 2000. A series of GEOs is intended for that service.
I am unaware of a third generation G* in 2003. I believe G* was interested in such in the year 2006. And I do not think pricing is an issue till 2003. That so many analysts think it will be before then reconfirms my belief that it will not be.
As for the companies you name: I would think LOR is looking to expand beamage, none of which the companies you list provide. They are more into networking and switching capacity.
Besides, a peice of GCOM is owned by Hughes, and Cable and Wireless owns a large percentage of AsiaSat.
Subject: Re: Probability of success
Date: Sat, Dec 27, 1997 20:44 EST
From: Readware
Message-id: <19971228014401.UAA23781@ladder01.news.aol.com>
Past is never prologue, but the launches G* is using have the following percentage success rates:
Delta II-- 95% (first 2 launches [4 satcoms each])
Zenit-2-- 88% (the 15th and 16th launch [the Tselinas] both resulted in explosion). Out of 28 launches 7 have mishapped, three resulted in permanent damage. The other four were eventually launched, but with time delays only. Zenit-2 has one more launch prior to the three launches of G*'s 12 satcoms/launch (the first is in the early summer). Out of 28 launches then, 25 have been without damage, four with time delays.
Soyuz: 97% success rate (I cannot find a failure in any of the manifests). These will launch 12 G* satcoms, four per launch (three launches).
Lockheed and SSL personnel will be launching the G* satcoms.
Averaging out the success rate: 5 launches should have a 96.3% success rate (Delta and Soyuz), three launches 88%. The launch rate weight is 91.7% for 8 launches. A comment on the Zenit-2 launches: Lockheed and SSL personnel did not launch these satellites where the failures were. Ground personnel makes a large difference in launch rate successes.
Subject: Re: Probability of success
Date: Sun, Dec 28, 1997 10:22 EST
From: Readware
Message-id: <19971228152201.KAA17147@ladder02.news.aol.com>
Probability theory does not work in predicting the success of any event whose outcome is not necessitated-- I am surprized you would would even cite it in the case of launch rates.
Indeed, some theoreticians hold no event is necessitated, and for that reason no event is predictable. It is not the case, in discussing a future event (of which probabily theory is one means of discourse), that the assertion of its occurrence is either true or false-- the assertion may be neither. Probability theory holds that the truth of excluded middle holds for every assertion, i.e., the assertion must be either true or false. (Probability
theory, by its very own terms, does not allow for an indeterminable outcome) That an assertion about a yet-to-be event can only be true or false has yet to be proven-- and it most likely never will be.
The other failing of probability theory is that it assumes reality can be mathematicized: mathematiization of reality equals total comprehension of it. Construct a one-to-one number series for each and every event such that the aggregate number series can substitute for all actual events of Nature-- and you will be able to divine the underlying algorithm, which divining gives you actual comprehension of Nature, and thus what events will take
place in it given what events have taken place in it.
I would think that Goedel's incompleteness theorem, wherein it is stated no mathematical system is self-validating, would also make one a bit reluctant in taking probability theory too seriously. And the Cartesian assumption that Nature, reality, can be mathematicized-- Newton's "God is a geometrician"-- is pretty much disregarded in modern day physics. Indeed, now "laws" of Nature are considered to be simply "approximations". And probably the
deathknell of all probability theory is quantum doctrine, wherein it is pretty well shown that no event has predictability.
Summing up: no launch rate success can be predicted ("past is not prologue", as I wrote). And probability theory-- which also, incidentally, includes a "closed set"-- is not really taken seriously in higher level physics. It can give no guidance in our search for certainty in areas whose outcome is not predetermined.
Subject: Re: Probability of success
Date: Sun, Dec 28, 1997 18:20 EST
From: Readware
Message-id: <19971228232000.SAA08684@ladder02.news.aol.com>
I did not say his math was not correct. All I said that is that mathematics is a totally helpless tool for predicting events that are not necessitated. That is why calculating the "probability" of a satellite launch's outcome is not an unguent to an investor: because mathematics (probability calculation) cannot bring him any closer to the knowledge of an actual outcome than simply not calculating the "probability" at all will do. Newton's
world has long been left in physics and mathematics. That is commonly understood, and simply reflects the common-sense notion that "Whatever will be, will be".
Some problems with "probability" theory (in layman's terms)-- if anyone is interested. It all depends on where you begin your calculation to come to a "probability" result. With what launch event do I begin to calculate the success of any launch? Sputnik 1, and the seven thousand launches from there? Or do I begin with Zenit-2's first 28 launches to determine the success of the 29th as far as G* is concerned?
And what variables do I include in the Zenit launch calculation-- only the days where the wind was 20 miles per hour at the time of the Zenit launches? Or where there were no palpable atmospheric pressures at all? We must include the probability of "calm winds" vs. "obnoxious winds" in our overall calculation of the probability of success of launch # 29, And do I include only GEO launches in Zenit-2? It has never launched a LEO satcom? And since it
has never launched a LEO satcom, where do I begin with calculating the probability of a G* LEO launch from the Zenit launch pad being successful? I cannot calculate it from a prior LEO Zenit launch, since there was none. Hence, for the Zenit-2 launches, I cannot calculate any "probability" of outcome for the G* LEO's first launch at all.
And do I include the Iridium LEO launches, and Pegasus LEO launches, in calculating G*'s LEO launch success rate? After all, I should completely eliminate all GEO launches, since a GEO launch has nothing to do with our concern here, namely the probability of success in a LEO launch.
And do I consider the personnel in the launching, as a variable, for determining the probability of success? Do I regress to a mean all applicable talents included and try to plot that on to the launch capabilities of Loral personnel? And on just how many of those variables do I regress? And how far do I extrapolate the past success of all launch team personnel-- both Loral and non-Loral on to a future launch that has only perhaps Loral personnel
staff launching?
And what about considering that no G* satcom part has any Hughes satellite parts in it? Does that figure into the variables that we must consider before calculating the probability of outcome?
Probability theory assumes mathematicization of an outcome can approximate the outcome more than not having done the probability calculation. And that is simply not the case. If the basic premis is that no event is necessitated, then probability theory is totally vacuous for predicting with any more cetainty than a "maybe yes, maybe no" statement.
Mathematicization of contingent events requires that all past variables possibly influential on a future event --all of them-- be factored into an equation that would probably be one billion miles long, one line. There is no way short of galactic omniscience where an individual could know what variable was applicable to a contingent event, and hence the outcome that contingent event may effect on a subsequent. As I stated Goedel's theorem shows the
inadequacy of all mathematical procedures as tools of certainty, and that goes for Boolean probability theory.
This is standard fare in doctoral courses. Nothing written here is new or grounbreaking. The future launch success rate of any satellite is totally unknowable. No amount of mathematicization (probability calculation) will get one any closer to the truth of the actual outcome. Unfortunate, but true. And no theoretician believes calculating probability of an outcome simply means taking a string of linearly successive events in the past (x1, x2, x3,
x4, x5...), without calculating all other applicable variables affecting those events, as Brian did, and then suddenly you have an outcome probability percentage that is meaningful for the future.
In satellite launches you want to manage risk-- and be able to deal with it should an unforeseen event (did I say unforeseen event?) occur. An experienced management is in whom investors place their money precisely for this reason. But if probability had any signifigance to an insurer, he certainly would not call satellites a "high risk" business. And the premiums they charge for satellite coverage certainly manifests their lacj of credibility in
"probability" theory.
Subject: Re: LORAL
Date: Mon, Dec 29, 1997 22:55 EST
From: Readware
Message-id: <19971230035501.WAA20456@ladder02.news.aol.com>
In June of 1997, Toothog, I wrote on this board that LOR, pending a successful first launch of G* in December, would reach $24/share. The target was based on a pricing model we use. It hit $24 1/4 before the launch delay was announced. With the launch delay, it went back down to the current price of $21.
Our pricing model is fairly predictive. When G* was at $27/share I suggested G* would reach $48/share in 1997. It hit $60 in 1997.
The pricing model has hit two-for-two.
Loral needs satellites in orbit to make money. That is what attracts investor interest-- revenues, then earnings.
With the G* launch in February, LOR will be $24. But the sats have to be in-orbit for the stock to move in price. The second launch will get LOR to $26.
On the need for satellites in-orbit:
In 1998, LOR will be launching Orion 2, Telstar 6, & 7. This is in addition to the already in-orbit GEOs Telstar 4, 5, and Morelos II, and Solidaridad 1 & 2. At the end of 1998, Loral GEOs will have 548 leased transponders.
In 1999, LOR will be launching Orion 3, Telstar 8&9, Cyberstar 1, and a replacement GEO for Satmex. At the end of 1999, its GEOs will have 674 leased transponders.
In 2000, LOR woll be launching Cyberstar 2 & 3. It will have 15 operating GEOs, with 780 leased transponders.
It is 1997: Currently LOR GEOs have 104 leased transponders.
Transponders equal revenues. The more transponders, the more earnings.
As some movie actor this past year said, "Show me the satellites".
Subject: Re: LORAL
Date: Tue, Dec 30, 1997 09:44 EST
From: Readware
Message-id: <19971230144401.JAA08090@ladder01.news.aol.com>
G*'s second launch? It is in early April from Cape Canaveral.
Subject: New Loral Orbital Slots
Date: Wed, Dec 31, 1997 14:09 EST
From: Readware
Message-id: <19971231190901.OAA29821@ladder01.news.aol.com>
The FCC yesterday awarded Loral 5 AT&T orbital slots, which AT&T surrendered upon their entrance from the satellite industry. Hughes was awarded the other 5.
Subject: Loral's "EBITDA Multiple"
Date: Thu, Jan 1, 1998 23:03 EST
From: Readware
Message-id: <19980102040300.XAA11360@ladder02.news.aol.com>
For three emails to me: what is the multiple one puts on Loral's EBITDA? That is, what multiple does one put on Loral's earnings before income taxes, depreciation, and amortization.
A multiple is the price you are willing to pay for a company's net earnings or its EBITDA.
A satellite company historically trades between 7 and 11 times its EBITDA.
In the case of Loral, it is not really just a satellite company. It is also a telecommunications company (with G*) and in 1999 too will be a full service Internet provider. A satellite multiple would be probably distinctly inappropriate for it then, to answer the question.
A stock normally trades on forward earnings.
If our earnings (EBITDA) forecast for Loral in 1998 are right (and I am very confident they are), Loral will earn $256.2 million. Based on 230 million shares outstanding, Loral is trading at roughly 21 times its forward (1998) EBITDA. (This has no G* revenues at all in it).
Assuming all the launches for G* in 1998 are successful, Loral in 1999 should earn $826 million (this number is lower than Loral's published pro forma estimates). On 280 million shares outstanding at the end of 1998, that would be an EBITDA of $2.95/share.
Do you give that a 21 multiple-- the mutliple it currently has? We do not. At the end of 1998 for a variety of backtesting reasons we believe Loral will be $36/share. However, if at the end of 1998 you give it the 21 multiple (it currently has) on forward EBITDA (1999's), Loral becomes a $62 dollar stock at the end of 1998.
For shareholders of Loral, it would be wonderful, of course. And I do not want to mislead anyone, but that is how its current multiple works on 1999 EBITDA. The math is rather straightforward.
I do think, the point is, that a 7-11 multiple on Loral in the questions is too low given that Loral is not really just a satellite company after 1998, and it is to after 1998 that the market is looking. In 1999, in fact, its satellite manufacturing division will only account for just under 13% of its EBITDA.
This is one of the problems in coming to an appropriate price of Loral as it goes forward. There is some substantial educated guesswork involved in deciphering tax loss carry forwards from acquisitions, amortization and depreciation schedules and how they increase margins on transponders, and on and on.
Quite frankly, it is easier to calculate a forward price for G*. But anyway the investor may see somewhat of the difficulty in pricing Loral. We did assume for a variety of backtesting reasons a 21 forward EBITDA multiple in stating that LOR would trade at $24 in 1997. It did trade slightly above there, though it closed at $21 7/16 for the year (because of the launch delay). We did say G* would close the year at $48, though it traded up to $60. It
did close the year at $49 and a fraction.
I think our $36 price is more realistic than the price a 21 multiple on 1999's EBITDA in 1998 would yield. We have a year to see. |
