The Destroyer of Moats

On a recent Cae­sar pod­cast I was talk­ing about the Roman siege against Masa­da in 73 C.E. and, of course, it remind­ed me of val­ue invest­ing.

If you don’t know the sto­ry, let me catch you up.

Scene: The First Jew­ish-Roman War.
Date: 66 C.E. — 73 C.E.

Judaea was under Roman con­trol when a Greek guy sac­ri­ficed a bird at the entrance of a syn­a­gogue in Cae­sarea Mar­iti­ma, a port city of the coast of Judaea and, at the time, the cap­i­tal of the province. That lead to protests, which in turn lead to some of the pop­u­la­tion refus­ing to pay their tax­es to Rome, which lead to the Roman gov­er­nor of the region going into the tem­ple to take the back tax­es, which lead to riots, and even­tu­al­ly a civ­il war between the Jews. A group of Jew­ish assas­sins, the Sicarii, so-called because of the large curved blade or “sica” they hid beneath their cloaks (the orig­i­nal ‘cloak and dag­ger’ oper­a­tion), mur­dered the High Priest along with thou­sands of their own peo­ple (men, women and chil­dren), and hun­dreds of Roman sol­diers, includ­ing the gar­ri­son at an impres­sive moun­tain-top fortress in the desert called Masa­da.

Fast for­ward to 73 C.E.

The Roman army, now lead by Titus, son of the new emper­or Ves­pasian, pret­ty much wiped out the Jew­ish forces, destroyed Jerusalem, burned their tem­ple to the ground, and then he sends one of his gen­er­als to clean up the Sicarii who are hold­ing out in Masa­da. The Sicarii are feel­ing pret­ty cozy. They are on top of a moun­tain fortress, with enough sup­plies to last them a cou­ple of years. No sweat.

But the Romans don’t take no for an answer. They start by build­ing an enor­mous wall around the moun­tain, so nobody can get in or out. Then they spend a few months build­ing a whop­ping great siege ramp, I’m talk­ing half a mil­lion tons of earth. On top of that, they build a humungous siege tow­er with a ginor­mous bat­ter­ing ram which starts pound­ing the wall. As the wall starts to crum­ble, the Sicarii, quick on their feet, start rein­forc­ing it… with dirt and wood. A clever engi­neer­ing solu­tion for with­stand­ing the impact of the bat­ter­ing ram. What do the Romans do? They sim­ply set fire to the wood.

By the time the Romans had breached the walls, the Sicarii had all com­mit­ted sui­cide.

Some­times a moat isn’t as pow­er­ful a defence as you think it is.

On this week’s Aus­tralian episode, we were talk­ing about REA — Reverse Engi­neer Any­thing, the AI frame­work that makes it easy to clone any appli­ca­tion that runs on your com­put­er. It will pull apart the code, fig­ure out how it works, and then build you the same func­tion­al­i­ty. And, of course, you can mod­i­fy it any­way you want, to suit your per­son­al cir­cum­stances. Soft­ware com­pa­nies that have spend years / decades build­ing up their prod­ucts may soon find them­selves los­ing cus­tomers who just build their own.

AFR had a sto­ry about soft­ware cloning last week:

United States software developer Brandon Thomas is behind the current interest in “cloned” products and announced ArtCraft last week. He used Anthropic tools to build versions that look like Adobe’s most popular apps, including Photoshop (PhotoCraft), Illustrator (VectorCraft), Acrobat Pro (PrintCraft) and Premiere (FilmCraft). He has since added knock-offs of Microsoft Word, PowerPoint and Excel. “Software is over … I just one-shotted Photoshop, the whole thing,” an account run by Thomas posted on message board Hacker News.

Check out Bran­don’s site.

Per­haps soft­ware moats are going to dis­ap­pear like Masada’s.

I know — cloning a piece of soft­ware isn’t the same as cloning an entire busi­ness. I worked at Microsoft long enough to know that there’s a lot more to the soft­ware busi­ness than just the prod­uct. How much of that will mat­ter, how­ev­er, when any­one can say to their AI “build me some soft­ware that does X, Y and Z”, I’m not sure.

What about tech­ni­cal sup­port? “Hey Chat­G­PT, fix that bug for me.”

What about new fea­tures? “Hey Claude, build me a new fea­ture.”

What about test­ing? “Hey Gem­i­ni, run a thou­sand vir­tu­al tests on the soft­ware, using every con­ceiv­able per­mu­ta­tion and fix any prob­lems you spot.”

Yes, Cameron, it’ll be ready in an hour.

Buf­fett, of course, has talked about “eco­nom­ic moats” for decades:

“The most impor­tant thing to me is fig­ur­ing out how big a moat there is around the busi­ness. What I love, of course, is a big cas­tle and a big moat with pira­nhas and croc­o­diles.” — “Strik­ing Out at Wall Street,” U.S. News & World Report, June 1994.

In QAV, we don’t have a score for “moat”. TK has actu­al­ly said “I’m not an investor who pays atten­tion to moats.” (#734)

But in an old episode when dis­cussing infla­tion, TK con­nect­ed moats and pric­ing pow­er direct­ly to what the QAV process cap­tures: “The process will still take that into account, and we’ll find stocks that can’t put their prices up will come off the buy list and those that can, will come on.” (#549)

In his pulled porks, he’s dis­cussed the rel­a­tive and per­ceived strengths of moats at com­pa­nies like Kel­ly Part­ners, JB Hi-Fi, Pro Medicus and IVE Group, and ques­tioned how much of one busi­ness­es like Kip McGrath and EROAD real­ly have.

If that com­pet­i­tive advan­tage helps a com­pa­ny main­tain its earn­ings and cash flow, we can pick up those results through the check­list with­out award­ing a sep­a­rate point for “moat”.

My point is that moats don’t last for­ev­er, as taxis found out when Uber arrived and Ency­clopae­dia Brit­tan­i­ca found out when Microsoft Encar­ta land­ed.

AI and humanoid robots are going to threat­en lots of moats over the next few years.

Which ones? How quick­ly? I have no idea. Some com­pa­nies will lose cus­tomers. Oth­ers will use the same tech­nol­o­gy to cut their costs and become even hard­er to com­pete with.

For­tu­nate­ly, I don’t have to work all of that out before I can invest. We can keep run­ning the check­list, buy­ing com­pa­nies that qual­i­fy, and fol­low­ing our sell­ing rules. That won’t get every deci­sion right. But it gives us a way to respond when things change, with­out hav­ing to pre­dict which soft­ware com­pa­ny will sur­vive the next clever girl with a lap­top.

And we need to be pre­pared to fol­low those rules even when we still love the busi­ness. A com­pa­ny can have a famous brand, a bril­liant prod­uct and decades of suc­cess behind it, and still become a poor invest­ment at the price we paid.

I’m hap­py to own shares in a com­pa­ny with a moat. I just don’t want to become so attached to the fortress that I refuse to leave.

---

QAV Myth Killers is a week­ly col­umn in the QAV newslet­ter, tak­ing apart a piece of
invest­ing con­ven­tion­al wis­dom. Read the series, or
get it by email every Fri­day.

Gen­er­al advice only. Space­craft Pub­lish­ing Pty Ltd
trad­ing as QAV is a Cor­po­rate Autho­rised Rep­re­sen­ta­tive (CAR 001292718) of MF & Co.
Asset Man­age­ment Pty Ltd (AFSL 520442). This is gen­er­al infor­ma­tion and does not take your
per­son­al cir­cum­stances into account. See our
dis­clo­sure.