Selecting investments based on covariance with the value of charities

Edit: af­ter do­ing some ba­sic es­ti­mates, I’ve found that the value of do­ing this is very low, gen­er­ally less than 100 ba­sis points: http://​​​​9of14il.png There­fore this strat­egy would only make any sense as a tiebreaker be­tween very similar in­vest­ments.


All of my in­vest­ments are split be­tween US stocks, US bonds, and EU stocks, and here I will ex­plain why. The in­vest­ment ideas I will provide here ap­ply to any effec­tive al­tru­ist in­di­vi­d­ual or or­ga­ni­za­tion in­vest­ing funds with the in­ten­tion of us­ing them for al­tru­ist pro­jects.


Fi­nan­cial ad­vice for the com­mon per­son of­ten sug­gests that peo­ple di­vert some of their in­vest­ments to in­ter­na­tional mar­kets in or­der to miti­gate do­mes­tic mar­ket risk. Emerg­ing mar­kets in par­tic­u­lar (like China, Brazil, and South Africa), are con­sid­ered to have value be­cause the risks in those mar­kets are not very well cor­re­lated with Western mar­kets. Fron­tier mar­kets (like Sri Lanka, Kenya and Pak­istan) are similar in this re­gard. The least-de­vel­oped coun­tries, like Uganda and Malawi, seem likely to be similar al­though they are not com­mon in­vest­ment tar­gets for sev­eral rea­sons. In­ter­na­tional stock in­dices gen­er­ally have a mix across de­vel­oped and emerg­ing mar­ket stocks.

The limited co­var­i­ance of in­ter­na­tional, emerg­ing and fron­tier stocks with the US/​EU/​Ja­pan mar­kets means that a typ­i­cal risk averse in­vestor can benefit, re­duc­ing their risk while pre­serv­ing ex­pected re­turns, by tak­ing a port­fo­lio of US stocks and re­al­lo­cat­ing some of it to for­eign mar­kets. They may also sac­ri­fice ex­pected re­turns in this way—even if they be­lieve that US stocks will out­perform emerg­ing mar­ket stocks, it may still be pru­dent for them to in­vest in some emerg­ing mar­ket stocks, be­cause it will re­duce the sever­ity of their ex­po­sure to the do­mes­tic mar­ket.

How­ever, it is com­monly ac­cepted by now that al­tru­ists should gen­er­ally be less fi­nan­cially risk averse than other peo­ple. This im­plies that we shouldn’t worry too much about di­ver­sifi­ca­tion, but only about ex­pected value. So if an al­tru­is­tic in­vestor thinks that US mar­kets will beat emerg­ing mar­kets, they should only in­vest in the US, and vice versa. If they per­ceived no sig­nifi­cant differ­ence it wouldn’t mat­ter much.

The ba­sic idea

If you are in­vest­ing money and plan to use it in the fu­ture to sup­port a char­i­ta­ble cause, there is an ad­di­tional con­sid­er­a­tion which is rele­vant to your choice of how to in­vest: the value of each dol­lar of your dona­tions, con­di­tional on your in­vest­ments be­ing suc­cess­ful. If you ex­pect that al­tru­is­tic pro­jects’ value will be cor­re­lated with a par­tic­u­lar set of fi­nan­cial in­stru­ments, then you should pre­fer in­vest­ing in those in­stru­ments since the po­ten­tial prof­its will be more valuable to you whereas your po­ten­tial losses are more likely to oc­cur in the sce­nar­ios where your money doesn’t have much use any­way. This means that, rather than max­i­miz­ing ex­pected fi­nan­cial re­turns, we should seek to find in­vest­ments which are co­var­i­ant with the value of our dona­tions.

The ba­sic idea, if you’re not con­vinced, is ex­plained in greater de­tail here.

Poverty donations

If you plan to make dona­tions to re­duce poverty, then you should seek to cor­re­late in­vest­ment re­turns with sce­nar­ios in which aid to the de­vel­op­ing world is more effec­tive. Now the rea­son that aid to the de­vel­op­ing world is so much more effec­tive than aid to the US or other coun­tries is that peo­ple there have much less in­come and there­fore have more ba­sic needs. So the ideal fi­nan­cial in­stru­ment for a poverty-fo­cused al­tru­ist who is sav­ing money would be some­thing that hedges against a rise in the wealth of the least de­vel­oped coun­tries. The most ob­vi­ous choice here would be to short-sell the stocks of com­pa­nies based in those na­tions. Now in­vest­ing in the least de­vel­oped coun­tries is already a very niche area of fi­nance, so I don’t know what the prospects are for short­ing those stocks. There may be other fi­nan­cial in­stru­ments which are suffi­ciently in­versely co­var­i­ant with the economies of the least de­vel­oped coun­tries to be worth pur­su­ing, but I can’t say for sure what they are.

This also re­lies on the as­sump­tion that a grow­ing econ­omy will also help the poor­est peo­ple in the coun­try. While this is not always strictly true, it’s a gen­eral trend of cap­i­tal­ist economies (“a ris­ing tide lifts all boats”). If you are con­sid­er­ing mak­ing these in­vest­ments, it may be worth in­ves­ti­gat­ing the de­gree and ra­pidity of this phe­nomenon, though even if the eco­nomic effects don’t play out quickly enough, a fis­cally stronger gov­ern­ment in an eco­nom­i­cally grow­ing na­tion will be able to di­rect its do­mes­tic health and welfare pro­vi­sion more suc­cess­fully, pro­vid­ing a similar effect.

An­i­mal rights and welfare

Meat con­sump­tion is strongly cor­re­lated with in­comes in de­vel­op­ing coun­tries, and so are “self ex­pres­sion” val­ues im­ply­ing a wider role for pro­gres­sive ethics in pub­lic life and de­ci­sion mak­ing. If you think that much of the value of an­i­mal ad­vo­cacy comes in the form of pre­vent­ing fu­ture meat con­sump­tion in cur­rently ris­ing economies, whether in emerg­ing, fron­tier, or least-de­vel­oped mar­kets, then you’ll want to be pos­i­tively ex­posed to those mar­kets: if these economies be­come strong, then peo­ple will ob­tain both the money to buy lots of meat and the so­cioe­co­nomic se­cu­rity to ex­pand their moral cir­cle, mak­ing ac­tivism more highly lev­er­aged. This means those coun­tries should get much or even all of your money.

How­ever if you think that the value of an­i­mal ad­vo­cacy is just in shift­ing Western prac­tices, then it’s not re­ally clear to me what to do. There is ev­i­dence that meat con­sump­tion even­tu­ally be­gins to re­verse with ad­vanced eco­nomic de­vel­op­ment, which would naively im­ply that you should hold short po­si­tions on U.S. mar­kets, but if chang­ing moral val­ues caused by eco­nomic de­vel­op­ment make ad­vo­cacy more effec­tive in the US (or if wealth makes peo­ple more open to buy­ing meat re­place­ments or hu­mane meat) then this may be the wrong po­si­tion to take. So whether to be short, long or in­differ­ent is not clear to me, but an­a­lyz­ing the is­sue may provide bet­ter an­swers.

An al­ter­na­tive ap­proach, in­stead of in­vest­ing on the ba­sis of mar­kets, is to make in­vest­ments ori­ented around com­pa­nies which spe­cial­ize in meat pro­duc­tion, an­i­mal feed, or meat re­place­ments.

Move­ment build­ing

If you view EA as some­thing that is go­ing to be good for set­ting global pri­ori­ties and grow­ing much larger than it is now, it makes sense to be pos­i­tively cor­re­lated with U.S., west­ern Euro­pean, and ANZAC mar­kets.

One rea­son for this is that lo­cal eco­nomic suc­cess will in­crease the amount of dona­tions which new EAs will be able to make, and pos­si­bly other fac­tors such as their ed­u­ca­tion and pro­duc­tivity. The great ma­jor­ity of our new mem­bers are from the West, and I be­lieve this is likely to con­tinue in­definitely due to lan­guage bar­ri­ers, ge­o­graphic dis­tance, and cul­tural val­ues. This means that the av­er­age in­comes of new mem­bers will be higher in the case that Western economies are strong, so re­cruit­ing new peo­ple will be higher in value. A good ob­jec­tion to this idea is that the wealthier EAs are and the larger the com­mu­nity, the more well-funded our pro­jects will be, so marginal fund­ing will be less im­por­tant. How­ever, the more peo­ple there are in the move­ment, the more new pro­jects we will have in dis­parate cause ar­eas, so the value of marginal dol­lars might not sig­nifi­cantly de­cline.

A sec­ond rea­son for cor­re­lat­ing move­ment build­ing funds with Western mar­kets is that it is im­por­tant for EA to be situ­ated in the cul­ture which has global dom­i­nance. It means that EAs will have ac­cess to the most pow­er­ful gov­ern­ments, the most in­fluen­tial in­di­vi­d­u­als, and the most elite in­sti­tu­tions of the world. Western coun­tries’ prac­tices and moral val­ues have spread to other coun­tries through a va­ri­ety of mechanisms due to our cul­tural, eco­nomic and mil­i­tary dom­i­nance of the global sys­tem (the former and lat­ter be­ing due at least in part to our eco­nomic dom­i­nance). If the West re­treats on its po­si­tion here, then EA will be less in­fluen­tial for the gen­eral course of global civ­i­liza­tion. But if the West re­mains dom­i­nant or be­comes even stronger rel­a­tive to other coun­tries, then EA will be more valuable.

A fi­nal fac­tor is that in­creased wealth in the West is likely to make peo­ple more re­cep­tive to ideas of al­tru­ism—see the world val­ues sur­vey linked above. This means that efforts in EA mar­ket­ing will see a higher rate of re­turn.

Ar­tifi­cial intelligence

I think that AI safety donors should be pos­i­tively cor­re­lated with U.S. AI com­pa­nies.

One is­sue here is the sec­ond rea­son de­scribed un­der move­ment build­ing. AI efforts in the U.S. are situ­ated in the pres­ence of re­search and ad­vo­cacy efforts where we have some mea­sure of in­fluence and lev­er­age. If com­pe­tent AI is de­vel­oped in China or Ja­pan then it will be less in­fluenced by our safety and ethics ad­vo­cacy than it would be if it were de­vel­oped in the U.S. It’s pos­si­ble that AI ethics and safety or­ga­ni­za­tions may arise in those coun­tries in the fu­ture, al­though we will be less able to com­mu­ni­cate with them and eval­u­ate the value of fund­ing them due to lan­guage and dis­tance bar­ri­ers, and they will be un­likely to match well with EA val­ues due to the lack of EAs over there along with cul­tural and moral differ­ences. This is more true if pow­er­ful AI is de­vel­oped with some level of in­sti­tu­tional se­crecy, as it is likely to be if strate­gic con­sid­er­a­tions are per­ceived by de­vel­op­ers or gov­ern­ments.

The rea­son this mat­ters here is that progress in gen­eral AI will lead to po­ten­tially enor­mous eco­nomic re­turns and it’s likely that these re­turns will be most con­cen­trated in the com­pa­nies which ac­tu­ally de­velop AI.

A sec­ond ar­gu­ment is that the US/​West is cur­rently the world leader in AI de­vel­op­ment, and game the­o­retic mod­el­ing im­plies that hav­ing a clear world leader in AI im­proves the de­gree to which safety con­sid­er­a­tions will be ad­hered to. So efforts for ad­vo­cacy and re­search in safety and ethics will make more head­way if the US re­mains in the lead than if com­pet­ing na­tions be­come com­pet­i­tive, as China is be­gin­ning to do.

A third ar­gu­ment is that even hold­ing con­stant the prospect of gen­eral AI be­ing de­vel­oped fastest in the US/​West, a more rapid de­vel­op­ment of AI in­creases the value of our re­search, since it re­duces the win­dow of time available for safety work and ad­vo­cacy.

Does eco­nomic growth re­ally mat­ter?

It may seem odd to claim that the fluc­tu­a­tions of mar­kets mat­ter much for these long-run con­sid­er­a­tions about the tra­jec­tory of the global sys­tem. How­ever, fluc­tu­a­tions over these timeframes aren’t just a re­flec­tion of pre­vi­ous growth in eco­nomic out­put; they also sig­nal in­vestors’ ra­tio­nal ex­pec­ta­tions about the value to be had in var­i­ous mar­kets.

A note of cau­tion: effi­cient mar­kets doesn’t im­ply equal ex­pected returns

Though you may ac­cept the Effi­cient Mar­kets Hy­poth­e­sis, you shouldn’t as­sume that any fi­nan­cial in­stru­ment where you have a per­sonal com­par­a­tive ad­van­tage in prefer­ring its var­i­ance is one that you ought to choose with­out do­ing some mea­sure of due dili­gence. Many in­vest­ments, such as fron­tier mar­ket stocks and any­thing used for hedg­ing, are se­lected for rea­sons which don’t fit well onto risk/​re­turn curves—namely, their limited or nega­tive var­i­ance with main­stream mar­kets. This means that an effi­cient mar­ket may in­vest in them to a de­gree that makes their ex­pected re­turns lower than that of other fi­nan­cial in­stru­ments with equal var­i­ance. So you shouldn’t as­sume that you should au­to­mat­i­cally make a cer­tain in­vest­ment just be­cause your al­tru­ism lines up with it; or­di­nary in­vestors of­ten have similar rea­sons for mak­ing those in­vest­ments.


For most cause ar­eas, there are ar­gu­ments that EAs with in­vest­ments which they plan to donate should al­lo­cate funds on the ba­sis of the ex­pected value of dona­tions con­di­tional on cer­tain in­vest­ments suc­ceed­ing or failing. The ar­gu­ments may be strong enough to com­pel one to in­vest some­where with a lower ex­pected fi­nan­cial value, or to make in­vest­ments where oth­er­wise one might have cho­sen to donate im­me­di­ately be­cause of haste con­sid­er­a­tions.