The Math of OlympusDAO

Keone Hon
19 min readJan 27, 2022


OlympusDAO, launched in March 2021, was arguably THE DeFi story of 2021, kicking off the DeFi 2.0 movement, pioneering the omnipresent (3,3) meme on Twitter, and generating fantastical returns for its early believers. Many people were fooled.

Its recent collapse (Jan 2022) may seem obvious in retrospect (“of course a 7000% APY was not sustainable”), but I believe most crypto-natives still don’t understand Olympus’ mechanics — what “backing” meant, where the high yield came from, or what elements of Olympus’ treasury were actually risk-free. These are important to understand in order to spot whether future treasury-backed protocols that we may encounter have merit.

The goal of this research piece is to elucidate the financial mechanics of OlympusDAO, introducing new nomenclature where necessary. I’ll try to answer: What powers the super high staking yield? To what extent is OHM “backed”? What’s up with Olympus’s claim that (3,3) is a dominant strategy?

All numbers (such as the OHM price of $413) in this piece are as of Dec 19, 2021. See the appendix for numbers taken from the OlympusDAO dashboard.

Please let me know of any errors via twitter. Above all, do your own research. This note does not constitute financial advice.

OHM aspires to become an algorithmic reserve currency backed by other decentralized assets. Similar to the idea of the gold standard, OHM provides free floating value that its users can always fall back on, simply because of the fractional treasury reserves where OHM draws its intrinsic value.

This note attempts to explain the mechanics of OlympusDAO. We will first walk through it in a simplified manner for ease of understanding, adding on some mechanical details at the end. Then we will discuss the dilutive effect that pOHM (an option held by insiders to mint OHM extremely cheaply) has on OHM’s treasury. Next we will discuss Olympus’s claim that it is resistant to a bank run. Finally we will discuss the innovations of bonding and protocol-owned liquidity.

A simplified explanation

The token of OlympusDAO is OHM. OHM is intended to be a “backed” currency, where the reserves in the OlympusDAO treasury are the backing for each OHM token.

The way a simple treasury-backed currency might work is: the treasury holds $50k worth of deposits and has 50k TreasuryCoins issued, and then when you bring in $1 worth of deposits, the treasury issues you 1 TreasuryCoin.

Like the hypothetical TreasuryCoin, an objective of Olympus is for each OHM to ultimately be backed by 1 DAI’s worth of direct treasury holdings. However, issuance of OHM is more complicated. Olympus allows OHM to be minted (“bonded”) at a floating rate based on a bonding curve; when this rate is significantly above 1 DAI, this means this mechanism will be bringing in excess DAI per OHM. This excess DAI will then be used to print more OHM, which will gradually be redistributed among OHM holders in the form of inflationary staking rewards. Newly bonded OHM is subject to a lockup that gradually vests over 5 days.

For example, the bonding curve might dictate that OHM can be bonded for 1.8 DAI. Then the extra 0.8 DAI could be used to back an extra 0.8 of OHM issuance, distributed to OHM stakers over time.

Why might OHM/DAI trade at a premium to $1? It could be because people are betting that more people will bond OHM at a later time. In that case it pays to bond OHM now to get access to later rewards. Or it could be because people are betting that OHM can eventually be worth more than the pure treasury value due to intangibles (mindshare, protocol additions, etc.). Whatever the reason, the premium generates profit when new people buy in via this bonding mechanism, which then powers benefits for existing holders.

As of Dec 19, 1 OHM can be bonded for ~400 DAI.

So, when you bond OHM using this mechanism, you bring in 400 DAI and OlympusDAO gives you 1 OHM and sets aside up to 399 OHM for everyone. That 399 OHM powers extremely high staking yields (in OHM terms) — currently 4500% APY. Given this exceptionally high yield, almost everyone stakes (unless they sell, more on that later).

The treasury holds $137M in DAI and other ordinary assets, which I’ll refer to as the “reserve risk-free value”, or RRFV, of the treasury. Meanwhile, there are currently 6.29M OHM circulating. Given no additional bonding, staking rewards will thus inflate the OHM supply 22x until it matches that 137M figure. Since everyone stakes, everyone’s OHM count will go up proportionally. Thus we can say that right now each OHM has a prorata claim on $22 of RRFV.

Protocol-owned liquidity

Although the above basic mechanic (pay huge premium to bond, then stake and receive portions of your and others’ premia) is powerful, it is further strengthened by a second bonding mechanism designed to ensure a liquid market for OHM.

The existence of a guaranteed liquid market for OHM helps instill greater confidence in OHM’s premium price. It’s not as scary to pay ~$400 per OHM when you see a deep market for OHM/DAI, and where the LPs have committed to never withdrawing. Olympus accomplishes this by allowing users to turn in OHM/DAI (and other OHM-related pair) LP tokens to bond OHM, also according to a bonding curve. This liquidity becomes locked permanently into the AMM, ensuring there will be liquidity on OHM pairs. Almost all of the liquidity on OHM-related pairs is owned by Olympus.

Olympus’s LP token holdings represent a claim on OHM and non-OHM assets. Olympus treats the claim on OHM as locked shares, which makes sense because selling your OHM back to the AMM for DAI is similar to burning OHM in exchange for the treasury’s DAI. Olympus computes the claim on non-OHM assets as part of its “backing” (e.g. it is part of “Market Value of Treasury Assets” on the dashboard). Let’s refer to the value of the non-OHM assets here as LPMV.

Olympus’s LPMV currently totals $559M, or $88.80 per outstanding OHM. LPMV serves as exit liquidity for people who decide to sell out of OHM, and it will fluctuate because net OHM buyers will increase it and net OHM sellers will decrease it.

How should we conservatively value something that fluctuates this much? A good lower bound is to value the LP tokens as if 1 OHM = 1 DAI, since Olympus’s policy is to continue diluting OHM until this is true. Under these circumstances, the ratios of OHM and DAI in the OHM/DAI pool will be 1:1, in which case both quantities will be sqrt(constantProduct), where constantProductis the current product of the current values in the pool. Thus, the value of all of Olympus's LP token holdings for a given pool are:

LPRFV = 2 * sqrt(constantProduct) * (% ownership of pool)

This allows us to define a new concept called total risk-free value (RFV):


Then Olympus policy for OHM issuance for stakers can be set using RFV rather than RRFV. That is to say, Olympus can print more OHM as long as supply is less than RFV.

You can see the explanation of the above formulas on Olympus’s docs.

Note that LPRFV ~= 2 * LPMV / sqrt(OHM_price) since LPRFV involves marking down our dollar assets by a factor of sqrt(OHM_price) in order to get the pool balances in the scenario when 1 OHM = 1 DAI. So (on Dec 19):

LPRFV = 559M * 2 / sqrt(413) = 55M
RFV = 55M + 137M
RFV/share = $30

The takeaway

A newbie, who has heard about OHM on twitter and wants to buy in, thus has the following options:

  • 1. Buy it on an AMM at the spot price (currently $413). Since all the AMM liquidity is provided by Olympus, you are effectively buying it from Olympus.
  • 2a. Purchase OHM via DAI bonds (“bond DAI for OHM”) by turning in about 400 DAI to Olympus for one OHM.
  • 2b. Purchase OHM via LP bonds (“bond LP tokens for OHM”) it by turning in LP tokens to Olympus as follows. Start with about 400 DAI, swap ~200 DAI for 0.5 OHM, post the 0.5 OHM and 200 DAI to the AMM, then turn in the LP tokens (a bit less than 1 OHM’s worth of LP tokens) to Olympus for 1 OHM.

Any of the above buying actions restocks the treasury’s resources at a price of ~400 DAI per OHM, Likewise, anyone selling OHM to the AMM will deplete LPMV at a price of ~400 DAI per OHM.

A simple mental model of buying OHM is that you are paying $400 for a claim on $22 of assets in RRFV and $5+ in LPRFV. You are also getting the value of OHM intangibles or future protocol revenue that Olympus could add on later, such as Olympus Pro (more on that later). Also if enough people buy in after you, the RRFV and LPRFV might go up.

pOHM’s dilutive effect

So far, we’ve come to understand OlympusDAO’s major holdings and its inflationary redistribution method. We have also examined the dynamics that might push OHM to trade at a premium to RFV, namely the prospect of future buyers and the interesting intangibles. Our mental model will be further complicated by the existence of pOHM.

pOHM is described in a post from March 13 on OlympusDAO’s Medium. (This post is also linked from the docs, and from OlympusDAO discord in response to the bot command ?pohm.)

Some of the important points from that post:

pOHM is a precursor derivative of OHM; it gives the holder the option to mint OHM by burning pOHM and providing the intrinsic value of OHM. For example, an investor would provide 1 DAI and 1 pOHM to mint 1 OHM.

This makes it similar to an option. pOHM is worth the price of OHM minus intrinsic value, and it only makes sense to redeem it when OHM is above intrinsic value. This ensures that our incentives are aligned to keep the premium alive. …

Team, investor, and advisor pOHM cumulatively vest as 11.8% of OHM supply. This means that at 1m OHM supply, a maximum of 118k pOHM can be redeemed. At 10m OHM supply, it’s 1.18m pOHM. pOHM holders finish vesting anywhere from 2b to 5b supply, so this is a long term bet. There’s a lot of upside for holders, but it is dependent on actual growth of the protocol.

The breakdown is as follows:

Team: 330m pOHM and 7.8% supply

Investors: 70m pOHM and 3% supply

Advisors: 50m pOHM and 1% supply

DAO: 550m pOHM and no supply cap (community can decide that!)

1 pOHM (previously called pOLY) is the option to mint 1 OHM for the cost of 1 DAI. In other words, it is the option to mint something currently trading for $413 for $1.

Early investors and the founding team hold 450m pOHM. Based on the article above, pOHM vests at 11.8% of the supply of OHM. Our estimates suggest 1% or 814k pOHM have been vested thus far. More pOHM will vest as more OHM is created, and OHM creation will happen naturally due to inflation.

pOHM dilutes OHM RFV. Although little of the total pOHM has vested so far, even the small amount that has vested is a substantial drag on Olympus’s ratio of marketcap to RFV.

It will be easier to reason about pOHM’s dilutive effect after inspecting a few scenarios:

  1. the current scenario
  2. a future scenario where marketcap remains constant relative to now, OHM supply has increased 4x due to scheduled staking emissions, and thus OHM price is 1/4 of what it is now:
# 1. Current scenario
6.3M OHM issued, OHM/DAI = 413, marketcap = 2.8B
740k pOHM vested
If exercised, OHM supply will increase from 6.3M to 7M
If exercised, RFV will increase from 192M to 193M
Value of newly created OHM position: 740k * 413 = 305M
Cost to exercise: 740k
# 2. Future scenario:
# Minimal bonding since now
# marketcap same as now
# OHM supply has increased 4x due to rebase
25.2M OHM issued, price = 103, marketcap = 2.8B
3M pOHM vested
If exercised, OHM supply will increase from 25.2M to 28.2M
If exercised, RFV will increase from 192M to 195M
Value of newly created OHM position: 3M * 103 = 309M
Cost to exercise: 3M

As you can see, pOHM is a call on 11.8% of the marketcap of OHM, with a variable strike price (in marketcap terms) which is 1/OHM_price of the notional value.

In OHM, bonders and open-market buyers currently pay a large premium over RFV; Olympus calls the excess beyond $1 “protocol revenue”, and it is redistributed to all OHM holders over time. pOHM adversely affects the economics for non-insiders because it gives insiders a sizeable claim on this premium.

Even if we attempt to compare this to typical governance tokens where insiders might account for 11.8% of the total supply, a notable difference still remains. In a normal situation with a fixed max supply of governance tokens, when insiders sell, their percentage ownership decreases and remains fixed. When insiders sell OHM, their collective percentage ownership drops, but then automatically begins increasing once more towards 11.8%. This is because insiders may own less than 11.8% of the OHM created up to now (due to the sale), but they will own 11.8% of the subsequent OHM created.

For example, consider the above pair of scenarios. In the current scenario, it seems that insiders could sell their 740k OHM into dollars for $305M (ignoring market impact!), at which point they would own 0% of the OHM supply… but then by the time that scenario 2 rolled around, they would still have 75% of what they had before (2.4M vs 3.2M), worth another 250M.

Commentators (such as the author of this thread linked from the ?threads command in Olympus’s discord) have tried to argue that insiders’ huge optionality via pOHM will cause them to optimize for long-term success. While this is true on a shallow level, the quirk of having insiders’ ownership converging back toward 11.8% even if they do sell amounts to a very generous insider allocation, especially for a protocol that heavily emphasizes not selling.


The bank run scenario

In the docs ( Olympus tries to directly address the scenario in which there is a run on the bank.

We think their assessment has it wrong:

What will happen if there is a bank run on Olympus? Fractional reserve banking works because depositors don’t withdraw their funds all at once. A depositor’s faith in the banking system rests on regulations and agencies like Federal Deposit Insurance Corporation (FDIC).

OHM does not have FDIC insurance but it has an incentive structure that protects stakers. Let’s take a look at how it performs during a hypothetical bank run. In this scenario, we assume the majority of stakers would panic and unstake their tokens from Olympus — the staking percentage which stands at 92% now quickly collapses to 3.3%, leaving only 55,000 OHM staked.

Next, we assume the Risk-Free Value (RFV) inflows to the treasury completely dry up. For context, RFV is currently growing at about $1 million every 2 days. However, during a bank run this growth will likely stop.

Finally, we assume that those last standing stakers bought in at a price of $500 per OHM. The initial investment of these stakers would be:

$500/OHM∗55,000 OHM=$27.5 million

As of September 15 2021, the total OHM supply is 2,082,553 and the RFV is $47,041,833. Remember that 1 OHM is backed by 1 USD (DAI or FRAX). By subtracting these two numbers, we know 44,959,280 OHM will eventually get issued to the remaining stakers. In roughly a year, these stakers who are holding 55,000 OHM will have:

55,000+44,959,280=45,014,280 OHM

$27.5 million investment made by these stakers will turn into about $45 million based on cash flow alone if they stay staked (recall that 1 OHM is backed by 1 USD). In this bank run scenario, the stakers who stay staked not only get their money back, but also make some profit. Therefore, (3,3) isn’t just a popular meme, it is actually a dominant strategy.

The above scenario is unlikely to play out because when other people find out that extremely high rewards are being paid to the stakers, they will copy the strategy by buying and staking OHM. This is also why the percentage of OHM staked in Olympus has consistently remained over 90% since launch.

What’s wrong here? The incorrect assumption here is that “the staking percentage collapses to 3.3%”, i.e. everyone else sells into the AMMs.

It would indeed be great for the 3.3% if 96.7% of the outstanding OHM cashed out into the AMM, since many of these OHM sellers would be selling at a super low price (well below RFV per share) in order for that to be the case.

The example above chooses an arbitrary percentage supply reduction such that the remaining shares would have an RFV per share of $900. 3.3% sounds like a super pessimistic scenario, but it’s actually super optimistic! In practice, in a bank run, we think stakers would cash out until the AMM presented them a price comparable to the prevailing RFV per share. This would reduce supply somewhat, which would raise RFV per share somewhat, but certainly not to above the current market price of $500.

Let’s simulate it more precisely!

In the above example (from Sep 15), RFV = 47M and OHM supply = 2.08M, so RFV per share is $23.
LPMV is not given in the example, but we’d estimate it to be ~4x the RFV (similar to what it is now), so $80 per share or 160M dollar equivalents.

Let E be the equilibrium RFV per share, and S be the equilibrium circulating supply. Let T be the amount of OHM and D be the amount of dollar equivalents in the AMM at equilibrium. Note that the AMMs (in aggregate, but we can treat them like a single AMM) start with 160M dollars and 320k OHM, so by the constant-product formula T * D = k where k = 160M * 320k = 5.12e13.

We have the following equations:

# final circ supply + final OHM in AMM = init circ supply + init OHM in AMM:
S + T = 2.08M + 320k = 2.4M # (1)
# RFV/shr * circ shares = RFV
E * S = 47M # (2)
# AMM constant-product is preserved:
D * T = 5.12e13 # (3)
# AMM price is E:
D / T = E # (4)

We have 4 equations and 4 variables, so we can solve fully.

E * (2.4M - T) = 47M  # from (1) and (2)
E = 5.12e13 / T^2 # from (3) and (4)
5.12e13 / T^2 * (2.4M - T) = 47M
multiplying both sides by T^2/47M:
1.089M * (2.4M - T) = T^2
T^2 + 1.089M*T - 2.612e12 = 0
solving for T:
T = 1.16M
solving for the others:
S = 2.4M - T = 1.24M
E = 47M / S = $37.9
D = E * T = 44M

So to summarize, in a bank run, about half of the circulating supply would sell into the AMM until the AMM had a prevailing price of $37.90, at which point the RFV per share would also be $37.90. The remaining half would stake to rebase proportionally, until enough OHM was issued to match the RFV (i.e. 46M more OHM was issued), at which point 1 OHM = 1 DAI.

In this bank run scenario, contrary to Olympus’s docs, the last standing stakers would be down over 90% from $500 to $37.90. It would be better to exit Olympus as early as possible during the bank run by selling into the AMM.

(Note, this analysis doesn’t even consider the additional dilution from pOHM which would unlock as OHM supply increases.)

OHM as a better memecoin

Let me attempt the following steelman argument:

DOGE has no intrinsic value but trades at a 27b marketcap. So where’s the problem with Olympus capturing a ~3b marketcap with only 190m book value? Perhaps the meme value of OHM is worth that.

Further, OHM is a “better” version of DOGE because it has marketcap appreciation built in: by default if no one buys or sells, DOGE price will stay flat and holders won’t gain anything (let’s ignore the slight inflation due to mining rewards!), whereas with OHM, by default the supply inflates, the AMM price stays flat, and everyone gains!

I can’t claim any expertise on valuing memes, and the “programmed marketcap appreciation” is certainly appealing. The rallying cry of (3,3) is a definitely a good example of a powerful meme. Many people identify as Ohmies.

I will say that there is a difference in transparency of meme status. DOGE, SHIB, SAMO, and co. are pretty clearly memes. They feature silly pictures and funny names. Meanwhile Olympus is described on and in more serious terms.

DOGE is serious business

Nobody would purchase DOGE without acknowledging on some level that they are buying into a joke. Meanwhile Olympus’ website uses phrases like:

  • “Sustainable Staking APY”
  • “Investment Protection”
  • “OHM is backed by an ever-growing, income-generating treasury.”
  • “We’ve created a currency that is able to constantly grow purchasing power despite market conditions.”

More serious tone, more room for negative surprises.

Innovation in bonding / protocol-owned liquidity

Having said the all of the above, we do believe there is some merit to the bonding mechanism and protocol-owned liquidity (POL) scheme pioneered by Olympus, in the following two senses.

First, it is valuable for protocols to sell some of their tokens into stablecoins (or more stable assets, like ETH). As discussed by Hasu in a recent article, protocols should probably sell some of their native token into dollars to pay their contributors, to fund liabilities from hacks, etc. We see the bonding mechanism as a means of gradually selling tokens into dollars. Selling gradually reduces market impact and is less likely to spook investors.

One could also argue that selling “bonds” is a potential marketing improvement over ordinary selling of coins, due to excitement over this new mechanism, although this may be temporary.

it’s bonding, not selling

Second, and relatedly, we do see value in protocol-owned liquidity. It is valuable for a protocol to have a guaranteed liquid market in its coin. An investor should be more comfortable buying, knowing that some market will always exist in the coin. Furthermore, there is significant price and volume reflexivity in crypto. Price reflexivity leads to high volatility which makes liquidity provision less enticing.

For example, if the price of token drops by 50%, a liquidity provider is better off removing their liquidity if they believe either of the following:

  1. they think the token will drop by another 50%
  2. they think the token will revert by rallying 100%.

The problem is further hurt by the fact that a token whose price approaches 0 will also inspire little volume, which further reduces the incentive to LP.

POL reduces the liquidity/price reflexivity seen in so many minor altcoins. The fact that it is done gradually via bonding helps too, since it allows the liquidity to scale up with the general interest in the coin.


OHM is framed as a “reserve” currency that is “backed” by assets held in its treasury. However, most of the “backing” is stablecoin in an AMM that will be depleted if a fraction of the outstanding supply of OHM is sold into the AMM.

The RFV (currently $30 vs OHM price of $413; 7.2%) is a better estimate of the “backing” of OHM. However, even this overstates the terminal value of 1 current OHM (i.e. including inflation rewards from staking), because pOHM holders hold the right to mint OHM for $1 of stablecoin, which will further dilute RFV per OHM.

OlympusDAO has innovated through its invention of the bonding meme and protocol-owned liquidity, both of which seem to offer some valuable functionality for DeFi protocol treasury and liquidity management.

Appendix: mechanical details

We simplified a few details in our explanation:

RFV Holdings

Technically, Olympus can hold more than just DAI in the reserve component of its treasury. Currently, it can hold a few other stablecoins (FRAX and LUSD) and some risk assets too (ETH, CVX, xSUSHI). We don’t think this changes our mental model much, as these risk assets are still substantially less risky than OHM, but we mention it here for completeness.


Bonded OHM vests linearly over 5 days, so bonders can start redeeming some of their OHM as soon as the next rebase happens (every 8 hours).

With Olympus V2 (released right around the time of this publication), OHM newly created through bonding automatically earns staking rewards, eliminating the risk that you overpay for bonds by buying at too little of a discount relative to the vesting period. We see this as a good simplification that eliminates some confusion.


Bonding is how Olympus acquires assets for its treasury. Bonding allows Olympus to incentivize users to bring assets into its treasury.

For a given asset that Olympus seeks to acquire, bond prices are set as follows:

bondPrice = 1 + premium
premium = debtRatio * BCV
debtRatio = bondsOutstanding / ohmSupply

BCV is a parameter set by the policy team for each asset, which effectively sets a target amount of that asset to acquire. However, individual users are the ones actually acquiring the asset. Also, a user who wants a particular asset to be more heavily held by Olympus could cause Olympus to acquire more of it than is specified by BCV, simply by continuing to buy bonds that bring in that particular asset even after the price becomes unfavorable. As such, Olympus is an interesting example of decentralized asset management.

The reports published by the policy team give a sense of what policy decisions are decided via governance (generally, enabling new assets to be bonded) and what policy decisions are made by executive decision with feedback from the community (generally, the actual BCV values).


The last official mention of pOHM that we can find is the article linked above. It was written on March 13, but to our knowledge it is still up-to-date, as its url is the response to the bot command ?pohmin the Olympus discord. Likewise, this thread is one of the threads linked by the bot command ?threads.

If you have any new information about pOHM mechanics, please let me know.

Olympus Pro

Olympus Pro is receiving a small fee for bonding services to other protocols. These fees are stored in whatever currency they were paid in; they currently sum to $515k, or about $0.07 per OHM right now.

Inferring RRFV, LPRFV, and LPMV from the dashboard

Recall that for clarity we defined a few sums:

  • RRFV: value of treasury reserves (i.e. the component that is not LP shares)
  • LPRFV: min value of LP shares if 1 OHM = 1 DAI
  • LPMV: market value of non-OHM side of LP shares

However Olympus (via its dashboard) exposes RFV , ‘Market Value of Treasury Assets’, and the OHM price.

It may be helpful to be able to infer RRFV , LPRFV , and LPMV from the figures provided by Olympus. We can do it by solving the following equations:

  • MarketValueOfTreasAssets = RRFV + LPMV
  • LPRFV = 2 * LPMV / sqrt(OHMprice)

For example, on Dec 19:

RFV = 192M
MVTA = 700M
OHMprice = 413
LPMV = 10.16 * LPRFV
RRFV + 10.16 * LPRF = 700M
RRFV = 137M
LPMV = 559M

Appendix 2: values from OlympusDAO’s dashboard

On Nov 25:

  • Market cap: $4.022B
  • OHM price: $814
  • Index: 38.5
  • Circ supply: 4,939,963
  • Total supply (this is circ supply + Olympus’s OHM on AMMs): 5,894,215
  • Backing per OHM (RRFV/share + LPMV/share): $163.41
  • Market value of treasury assets (RRFV + LPMV): $805M
  • RFV: $178M
  • RRFV: $130M
  • LPRFV: $47M
  • LPMV: $675M
  • Total value deposited (OHM price * shares staked) $3.609B
  • OHM staked: 90.07%
  • APY: 7116%

On Dec 19:

  • Market cap: $2.77B
  • OHM price: $413
  • Index: 49.39
  • Circ supply: 6,291,417
  • Total supply: 7,055,595
  • Backing per OHM (RRFV/share + LPMV/share): $111.60
  • Market value of treasury assets (RRFV + LPMV): $702M
  • RFV: $192M
  • RRFV: $137M
  • LPRFV: $55M
  • LPMV: $559M
  • Total value deposited: $2.539B
  • OHM staked: 91.43%
  • APY: 5261%

Further reading

Thanks to Jordi Alexander, Maher Latif, and Jayant Krishnamurthy for feedback on this piece.